Epigenetic Fingerprints Link Early-Onset Colon and Rectal Cancer to Pesticide Exposure

Epigenetic Fingerprints Link Early-Onset Colon and Rectal Cancer to Pesticide Exposure | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Analysis Epigenetic Fingerprints Link Early-Onset Colon and Rectal Cancer to Pesticide Exposure Silvana Maas, Iosune Baraibar, Odei Blanco-Irazuegui, Josep Tabernero, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4528579/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 21 Apr, 2026 Read the published version in Nature Medicine → Version 1 posted You are reading this latest preprint version Abstract The incidence of colorectal cancer (CRC) is rapidly rising in individuals younger than 50, particularly in high-income countries. This rise parallels shifts in lifestyle and environmental factors, collectively termed the exposome; however, whether these are causally linked to the development of early-onset CRC (EOCRC) has not been investigated. Due to limited exposome data in most cancer cohorts, we constructed weighted methylation risk scores (MRS) as proxies for exposome exposure to pinpoint specific risk factors associated with EOCRC. Our analysis confirms previously identified risk factors, such as educational attainment, diet, and smoking habits. Moreover, we identified the exposure to the herbicide picloram as a novel risk factor (Padj. = 0.00049), a result we replicated in a meta-analysis comprising six CRC cohorts (P = 0.021), comparing EOCRC cases with patients diagnosed aged ≥70. Subsequently, we employed population-based data from 81 U.S. counties over 20 years and validated the association between picloram usage and EOCRC incidence (P = 2.87×10 -3 ). These findings highlight the critical role of the exposome in EOCRC risk, underscoring the urgency for targeted personal and policy-level interventions. Biological sciences/Cancer/Gastrointestinal cancer/Colorectal cancer Biological sciences/Computational biology and bioinformatics/Data integration Health sciences/Risk factors Biological sciences/Molecular biology/Epigenetics/DNA methylation Biological sciences/Cancer/Cancer epidemiology Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Globally, colorectal cancer (CRC) ranks as the third leading cancer type and the second most common cause of cancer-related death ( 1 ). As CRC is an aging-associated disease, the rates of CRC incidence and death grow steadily with age, with an estimated 90% of worldwide cases and deaths occurring in individuals over 50 years old ( 2 ). However, in recent years, worldwide cancer registries have reported a disproportionate uptick in the incidence of early-onset CRC (EOCRC), generally defined as CRC diagnosed in individuals younger than 50 years of age ( 3 , 4 ). In the U.S., the average annual percent change in the incidence of EOCRC was 2.2% between 2008 and 2012, while similar trends have been observed in New Zealand, Canada, and Europe ( 3 , 5 ). This tendency is expected to continue in the coming years, with a predicted increase in the incidence of rectal and colon cancer in patients below the age of 34 by 2030 of 124% and 90%, respectively ( 6 ). EOCRC presents unique clinical and pathological characteristics in comparison with CRC in older patients, with a predominance of rectal and left colon tumors, higher prevalence of synchronous and metachronous CRC, higher frequency of metastatic disease at diagnosis, and more aggressive and less differentiated tumors ( 7 – 10 ). However, at the genome level, molecular alterations are similar to those described in older patients, and no specific alterations have been described ( 11 – 15 ). Concerning epigenetic changes, hypomethylation of long-interspersed nucleotide elements-1 (LINE-1), which acts as a surrogate marker of global DNA methylation status in the genome, has been reported in patients with EOCRC ( 16 – 18 ), while an assortment of lifestyle and environmental exposures have been associated with lower LINE-1 methylation levels in blood cells in the general population ( 19 – 21 ). As our understanding of risk factors for CRC expands, it is becoming increasingly evident that lifestyle factors, such as diet and exercise, can significantly influence CRC risk. Despite this growing knowledge, the reasons behind the rising incidence of CRC in young patients remain unclear. Cancer predisposition syndromes are found to be the underlying cause in only 10–30% of patients with EOCRC, hence most patients do not present any genetic predisposition ( 22 , 23 ). Emerging hypotheses suggest that lifestyle and environmental exposures — collectively known as the exposome — may play a critical role. However, efforts to identify modifiable risk factors specific to EOCRC have met with only partial success ( 24 ). Most studies compare early-onset patients with controls, primarily identifying the same risk factors as those for later-onset cases ( 25 ). The main investigated modifiable traits that are associated with EOCRC are smoking habits, diet, educational attainment, Body Mass Index (BMI), and microbiome ( 26 ). Concerning microbiome, patients with CRC present a distinct microbiome composition in comparison with healthy subjects. Although specific differences in the microbiome between EOCRC and those diagnosed at later ages remain unexplored, E. coli strains harboring the polyketide synthase (pks) operon, producing the genotoxic compound colibactin and described as contributors to CRC, have been found in CRC cases diagnosed at a younger age ( 26 , 27 ). One possible reason for this limited success is the lack of quantitative measurements of exposome traits in cancer cohorts, making it challenging to establish clear associations with cancer risk. However, emerging studies have identified exposome-induced changes in CpG site-specific DNA methylation levels ( 28 – 43 ), providing a new avenue for investigation. As DNA methylation data is often available in cancer cohorts, utilizing exposome-related methylation changes could serve as a valuable proxy for direct exposome measurements. This approach may enhance our ability to pinpoint specific exposome risk factors and improve our understanding of their role in EOCRC development. In this study, we aimed to explore the exposome traits that could be contributing to the development of EOCRC compared to later-onset CRC (LOCRC), here defined as CRC diagnosed in individuals 70 years and older, using epigenetic markers. To this end, we constructed methylation risk scores (MRS) for lifestyle and environmental factors using DNA methylation data from The Cancer Genome Atlas (TCGA) as discovery and six replication datasets and compared the constructed MRSs between EOCRC and LOCRC. Furthermore, we investigated the use of single base substitution (SBS) mutational signature to account for tumor age rather than the patients’ age at diagnosis. The relationship between pesticide use intensity and EOCRC incidence in the U.S. was further investigated using population-based data. Results Cohort Description For the discovery phase of this study, we utilized colon adenocarcinoma (COAD) samples obtained from TCGA (44). The subsequent replication phase was carried out through a meta-analysis, incorporating data from studies on colon cancer (GSE131013 (45) and GSE42752 (46)), rectal cancer (TCGA-READ and GSE39958 (47)), and colorectal cancer (GSE101764 (48) and GSE77954 (49)). We categorized the participants by age: those diagnosed before the age of 50 were identified as early-onset cases, while those diagnosed at 70 or older were designated as later-onset CRC (LOCRC) cases. The population characteristics across all datasets are presented in Table 1. Table 1: Study population Characteristics Study N total Females, N (%) Age, years mean (SD) Cancer type Discovery dataset: TCGA-COAD Colon Early-onset 31 18 (58) 43 (5) Later-onset 100 43 (43) 78 (5) Replication datasets: TCGA-READ Rectal Early-onset 14 6 (43) 44 (6) Later-onset 30 21 (70) 76 (6) GSE39958 Rectal Early-onset 12 3 (25) 45 (7) Later-onset 9 2 (22) 74 (5) GSE42752 Colon Early-onset 4 3 (75) 46 (3) Later-onset 7 5 (71) 76 (4) GSE77954 * Colorectal Early-onset 3 3 (100) 48 (1) Later-onset 10 3 (30) 77 (6) GSE101764 Colorectal Early-onset 13 3 (23) 38 (8) Later-onset 37 13 (35) 76 (5) GSE131013 * Colon Early-onset 2 0 (0) 46 (4) Later-onset 58 13 (22) 76 (5) Early-onset cases are defined as participants younger than 50 years, while later-onset cases comprise participants diagnosed at age 70 years and older. Dataset GSE77954 encompasses primary (n=7) and metastatic samples (n=6). * Notably, the early-onset category in GSE131013 comprises solely male participants while GSE77954 includes only female participants. Consequently, adjustment for sex was not feasible in these two datasets. Exposome-related DNA methylation marker sets In exploring the exposome's influence on early-onset versus later-onset colon and rectal cancers, our research focused on a curated list of 29 (28–43) lifestyle and environmental factors. The analyzed traits encompassed 10 lifestyle factors: the Alternative Healthy Eating Index (AHEI) (28), alcohol consumption (29), birth weight (30), BMI (continuous variable in kg/m 2 ) (31), coffee consumption (32), education level (33), Mediterranean Diet Score (MDS) (28), obesity (defined as ≥30 kg/m 2 ) (34), smoking habits (35), and smoking inference model (smoking-Maas) (36). Furthermore, we examined 5 air pollution particles: nitrogen dioxide (NO 2 ) (37–39), polychlorinated biphenyls (PCBs) (42), and particulate matter (PM) <10 micrometers (µm) in diameter (PM 10 ) (38–40), <2.5 µm (PM 2.5 ) (37–41), and between 2.5 and 10 µm (PM 2.5-10 ) (40). In addition, we included 14 pesticides encompassing 2,4-dichlorophenoxyacetic acid (2,4-D), atrazine, acetochlor, chlordane, dicamba, malathion, Dichlorodiphenyltrichloroethane (DDT), heptachlor, lindane, glyphosate, mesotrione, metolachlor, picloram, and toxaphene (43). For the marker selection, we identified for each trait significantly associated CpG sites from extensive epigenome-wide association studies (EWAS), employing various significance thresholds, namely P<1.2×10 -7 , P<1.0×10 -5 , and false discovery rates (FDR) of <0.01, <0.05, and <0.1. Exposome-related methylation risk scores Utilizing available EWAS summary statistics of each trait for the five marker selection thresholds, we identified 73 exposome CpG sets across the 29 exposome traits. These sets were used to compute 73 weighted methylation risk scores (MRSs), utilizing DNA methylation beta-values adjusted for epigenetic age estimators derived from the Horvath clock (50). The number of CpG sites associated with each trait and their respective significance thresholds can be found in Supplementary Table S1 and their corresponding weights in Supplementary Table S2. To elucidate the exposome impact on early-onset colon and rectal cancer cases, we compared the 73 MRSs across early- vs. later-onset (reference group) patients, using multivariate logistic regression models. As there are sex disparities in CRC incidence (51), we adjusted the regressions for sex, when possible. In the discovery dataset, positive associations were observed for MRSs related to PCB, PM 10 , the smoking-Maas model, heptachlor, metolachlor, picloram, and toxaphene, while negative associations were found for MRSs corresponding to BMI, education level, MDS, obesity, atrazine, malathion, and mesotrione (Fig. 1 and Supplementary Table S3). We highlight the data for four lifestyle factors previously linked to colon and rectal cancers, including the MDS (52) (Fig. 2a) and education level (53) (Fig. 2b), which are considered protective factors, as well as smoking habits (54) (Fig. 2c) and obesity status (55) (Fig. 2d), which are recognized as risk factors. To elucidate the directionality of our findings, the heatmaps in the left panels of Fig. 2 show the methylation level distributions across CpGs featured in each of the four MRSs, their direction in the original EWAS and sorted by the derived MRSs. The heatmaps depict that an increased MRS correlates with higher beta-values in CpGs with positive associations in the EWAS, and lower beta-values in CpGs with negative associations (please refer to Supplementary Fig. S1 for a more detailed explanation). These results suggest that an elevated MRS mirrors greater exposure levels in the original EWAS. Specifically, for patients with early-onset colon cancer, this suggests deviations from the MDS (Padj. = 0.037) (Fig. 2a), lower education levels (Padj. = 0.025) (Fig. 2b), increased smoking exposure (Padj. = 0.010) (Fig. 2c), and lower obesity rates (Padj. = 0.011) (Fig. 2d) in comparison to those with later-onset, as illustrated in the middle panels of Fig. 2. The association of lower obesity rates in early-onset cases was verified utilizing physical metrics from TCGA-COAD. Colon cancer patients with a BMI over 30kg/m 2 , as measured in the clinic, were categorized as obese, resulting in 4 out of 24 early-onset and 18 out of 72 later-onset patients marked as obese. This provides a relative risk (RR) of 0.67 (95% CI: 0.26-1.76) for obesity in early-onset colon cancer patients within the TCGA-COAD cohort (Supplementary Fig. S2), supporting the results obtained with the MRS for obesity, and validating our methodology. Extending our investigation, we conducted a meta-analysis of the six replication datasets (Fig. 1 and Fig. 2 right panels, and Supplementary Table S4). This analysis corroborated the initial findings, notably the associations of non-adherence to MDS (P = 0.011, Padj. = 0.051), lower educational levels (P = 0.0039, Padj. = 0.025), and higher smoking exposure (P = 0.0025, Padj. = 0.024) in EOCRC (Fig. 2, right panel). Moreover, we conducted separate meta-analyses of the datasets comprising only rectal cancer samples (TCGA-READ and GSE39958) and only colon cancer samples (GSE131013 and GSE42752). The results obtained in TCGA-READ and the meta-analyses are presented in Supplementary Fig. S3 and Supplementary Table S5. Picloram-related methylation risk scores Our results highlight a novel association between the MRSs for pesticide picloram and the incidence of early-onset colon and rectal cancer, in comparison to later-onset cases, in both the discovery and meta-analysis (Fig. 1). We explored the directionality of these results further, highlighting the data for the MRS employing the genome-wide marker selection threshold (MRS-GW). We observed that a higher exposure level, as indicated by the original EWAS direction, is associated with an elevated MRS (Fig. 3a). This association highlights an augmented exposure to picloram among patients with early-onset colon cancer (Padj. = 0.00049) (Fig. 3b), a finding consistently supported by our meta-analysis (P = 0.021; Padj. = 0.081; OR: 1.6 [95% CI: 1.07-2.38]) (Fig. 3c). To ascertain the reliability of our findings, we executed two distinct permutation tests to ensure that the observed associations stem from biological relationships rather than being artifacts of particular CpG selections or patient classifications. Initially, the examination of the CpG sites in our MRSs revealed that the CpG sites from the picloram MRS-GW were 13 th in terms of significance among 10,000 permutations (Fig. 3d). Furthermore, patient classification permutation identified age-based classification as the second most significant, based on picloram MRS-GW, among 1,000 permutations, as depicted in Fig. 3d. The outcomes of the CpG site permutations for MDS, education levels, the smoking-Maas model, and obesity are detailed in Supplementary Fig. S4a, while the findings from the onset categorization permutations are shown in Supplementary Fig. S4b. Young tumors associated with picloram exposure Current patient classification of CRC into early-onset or later-onset categories relies on the patient's age at diagnosis. However, this method is flawed, as the interval between tumor initiation and diagnosis varies significantly among patients, rendering age at diagnosis an unreliable indicator of the tumor's actual age. To address this, we assessed if the single-base substitution signature 1 (SBS1) score, an indicator of the number of mitotic divisions a cell has undergone (56,57), can instead be used as tumor age. For this purpose, we selected patients from TCGA-COAD with data available on DNA methylation and mutational signatures. Furthermore, we excluded patients exhibiting microsatellite instability (MSI), considering MSI arises from defective DNA mismatch repair, inducing distinct mutational patterns that might drive tumorigenesis through mechanisms different from those in microsatellite stable (MSS) tumors (Supplementary Fig. S5) (58,59). The distribution of SBS1 mutations across different age groups— early-onset, middle-onset (aged between 50 and 69 years), and later-onset— among the 173 patients included in the study is detailed in Fig. 3e. Upon comparing the early-onset cases (N = 25) against later-onset cases (N = 72) and the picloram MRS-GW, we observed a statistically significant difference (OR: 2.99 [95% CI: 1.70-5.85]; P = 4.27×10 -4 ). Next, we employed a SBS1 score threshold, identifying 72 young (SBS1<60) and 101 old tumors (SBS1≥60). The novel patient categorization underscored a significant association with picloram MRS-GW (OR: 1.84 [95% CI: 1.31- 2.66]; P = 6.57×10 -4 ) (Fig. 3f). The chronological age distribution for these SBS1-categorized young and old tumors is provided in Fig. 3g. Pesticide use and EOCRC incidence in population data The use of MRSs as a proxy for pesticide exposure identified significant associations with several pesticides. Next, we aimed to validate the obtained results employing pesticide use intensity and incidence of EOCRC for available overlapping counties in California, Connecticut, Georgia, Iowa, New Mexico, Utah, and Washington in the United States. Based on data availability, we extracted pesticide usage from the Pesticide National Synthesis Project encompassing acetochlor, 2,4-D, atrazine, dicamba, glyphosate, mesotrione, and picloram. The EOCRC incidence rates were extracted from the Surveillance, Epidemiology, and End Results (SEER), encompassing EOCRC rates measured in 8 registries from 1975 to 2020 (SEER8) or in 12 registries measured from 1992 to 2020 (SEER12). The total number of included observations, the number of measured years times the number of overlapping counties, for acetochlor (SEER8: N = 1.111, SEER12: N = 1.196), 2,4-D (N = 1.983, N = 2.059), atrazine (N = 1.871, N = 1890), dicamba (N = 1.909, N = 1.964), glyphosate (N = 2.002, N = 2.097), mesotrione (N = 636, N = 636), and picloram (N = 1.531, N = 1.548) are depicted in Fig. 4a. Specifically, Fig. 4b shows the average (log) picloram use intensity and the average EOCRC incidence rates between 1992 and 2012 in the state of Iowa. To assess the relationship between pesticide-use intensity (exposure) and age-adjusted EOCRC incidence rates, we utilized linear mixed models adjusting for the years of data collection and a random effect to accommodate county-level variations. Our approach also tested for interaction effects between pesticide use intensity and the years of data collection, which proved to be non-significant for all pesticides under study (data not shown). Significant associations were found between the pesticide use intensity of multiple pesticides and EOCRC incidence in both SEER8 and SEER12, including glyphosate (SEER 8; P = 1.18×10 -5 , SEER 12; P = 2.02×10 -4 ), atrazine (P = 1.81×10 -4 , P = 4.21×10 -3 ), picloram (P = 2.87×10 -3 , P = 1.82×10 -2 ), 2,4-D (P = 4.16×10 -3 , P = 1.8×10 -3 ), and dicamba (P = 4.94×10 -3 , P = 4.10×10 -2 ), depicted in Fig. 4c. Discussion In this study, we explored the exposome traits that could be specifically contributing to EOCRC disease through the use of epigenetic markers. There is growing evidence of a worldwide rise in the incidence of CRC cases in individuals younger than 50, while the number of cases in later ages is decreasing in most countries ( 60 ). Epidemiological studies have linked CRC to non-modifiable factors, including sex, age, ethnicity, and genetic predisposition ( 51 , 61 , 62 ), and to external exposures such as adiposity, smoking and drinking status, household income, Western pattern diet, and physical inactivity ( 63 – 71 ). However, these exposures have not been specifically associated with EOCRC, leaving the reasons for the rising EOCRC incidence rates unclear. To the best of our knowledge, this is the first research addressing the exposome impact — including environmental factors and lifestyle — on early-onset colon and rectal cancer cases through epigenetic fingerprints. The effects of environmental exposures such as microplastics and nanoplastics (MNPs), air pollutants, and other potential carcinogenic agents on human health and CRC are just beginning to be documented ( 70 – 77 ). Exposures through lifespan may change, so the analyses in terms of both categorical and quantitative measurement are challenging and may raise concerns about potential spurious associations. Our results consistently link EOCRC to exposure to the pesticide picloram based on epigenetic fingerprints. Picloram’s mechanism of action as an herbicide relies on its capacity to mimic the plant growth hormones auxins and to inhibit the enzymes that break down auxins, which leads to more persistent effects than the natural hormone ( 78 ). Therefore, picloram disrupts normal growth, causing abnormal stimulation and maturation of tissues, which triggers growth discontinuation, root deterioration and eventually plant death. Picloram was first registered as a pesticide in the U.S. in 1964 and the herbicide and its derivatives have generally shown to be of moderate to low acute toxicity in laboratory animals. However, dietary exposure to residues of picloram is plausible as it has been found in grain and meat by-products, and the effects of long-term use on human health have not been described so far. If the use of picloram in crops started in the mid and late 20th century, then-current individuals with LOCRC were not exposed to it during their childhood, while cases of EOCRC were and have been for a longer part of their lives, which could explain our results. The association between pesticide exposure and EOCRC using population-based data further validates the association with picloram. Besides picloram, we also show evidence for associations between EOCRC and exposure to glyphosate and atrazine. Glyphosate is already categorized as “probably carcinogenic to humans” by the International Agency for Research on Cancer (IARC), suggesting the validity of the obtained results ( 79 ). Although atrazine has been banned in the European Union since 2004, the U.S. EPA still approves its continued use ( 80 ). Already during the IARC review in 1999, sufficient evidence confirmed the carcinogenicity of atrazine in experimental animals, but as evidence showed that the identified mechanism in rats was not relevant to humans, atrazine was categorized in group 3 (“not classifiable as to its carcinogenicity in humans”). Novel studies have identified alternative mechanisms for how atrazine could cause carcinogenesis in humans, e.g. by damaging DNA integrity, the stability of the cell genome, DNA double-strand breaks, and the activation of DNA damage checkpoints ( 81 ). This resulted in the Monographs program for 2025–29 ( 82 ) to warrant a high-priority re-evaluation of the IACR classification as a result of the identified new human cancer and mechanistic evidence. Furthermore, a recent update study in the Agricultural Health Study Cohort further suggested the link between atrazine use and several cancer types, including in patients < 50 years of age ( 83 ). Lower education levels, increased smoking exposure, non-adherence to MDS, and lower obesity rates for patients with EOCRC were found. This suggests that obesity could be a greater contributing factor in LOCRC than in EOCRC. Obesity has been largely described as a risk factor for CRC, although contradictions about its role in tumor aggressiveness, CRC progression, and survival have been reported. For EOCRC, a meta-analysis showed that a BMI greater than 30 kg/m 2 was significantly associated with the development of the disease ( 84 ) as compared to controls, but considerable heterogeneity among risk estimates was found. Conversely, when comparing EOCRC with patients diagnosed at a later age, older patients were more likely to be obese or overweight ( 85 – 87 ). Of note, in our study, we lack data about longitudinal BMI throughout the life course, but longer exposure through years to high BMI and adiposity in patients with LOCRC could be mediating this effect. It should also be considered that weight loss is a common symptom of CRC, especially in patients with advanced disease, which is, in general, more frequent in EOCRC. Further studies including longitudinal follow up are needed to confirm the role of obesity in EOCRC. The patient's age at diagnosis is currently used as the tumor's age, possibly introducing bias in identifying early-onset patients. To address this, we assessed if tumor age can be established by identifying molecular characteristics of aged tumors, such as mutational signatures ( 56 , 57 ). Characteristic mutational signatures in cancer genomes arise from various mutational processes, including defects in DNA maintenance mechanisms and both external and internal exposures ( 57 ). Mutations attributed to the single-base substitution signature 1 (SBS1) are thought to occur during DNA replication in mitosis, suggesting that the rate of SBS1 mutations may serve as an indicator of the number of mitotic divisions a cell has undergone ( 56 , 57 ). Furthermore, a correlation exists between the number of SBS1-attributable mutations within a tumor and the patient's age at cancer diagnosis ( 55 , 56 ). Here, we demonstrate that utilizing an SBS1 score threshold to distinguish between younger and older tumors is associated with picloram MRS. This finding suggests that the differences observed may be attributable to the biological age of the tumor rather than the chronological age of the patient. We acknowledge the limitations of the current study. First, the number of EOCRC cases in the datasets is relatively small. Second, we recognize low ethnical diversity in the investigation, as TCGA is mostly enriched in non-Hispanic white ethnicity. Although the other datasets consist of patients from Russia, Spain, Germany, South Korea, and the United States, independent studies on other populations are highly encouraged to capture the diversity of worldwide patients with EOCRC in terms of ethnicity, cultural, and socioeconomic differences. The study also presents unprecedented strengths. First, the use of MRS as a proxy for exposome factors is a novel approach that allows the exploration of traits that would otherwise not be possible due to limited data availability. Similarly, the lack of longitudinal follow-up measurements or the use of self-reported data can be overcome as MRSs integrate the cumulative impact of the exposure over time. Nevertheless, prospective birth cohorts with long-term follow-up, quantitative exposure measurement, and biomarker and omics analyses throughout life can elucidate the etiology of EOCRC. Plus, given that exposures in early life may be key for the development of CRC, prospective birth cohorts are needed. However, inherent limitations of these studies, such as cost, cohort size, healthy volunteer bias, and long follow-up needed for cases to appear and draw conclusions, should also be considered. Second, several independent cohorts were used, permutation tests were implemented, and the picloram MRS was consistently associated with EOCRC compared to LOCRC. Also, the validation of the association between picloram use intensity and EOCRC in population-level data further strengthens the evidence of picloram as a novel identified risk factor. Third, the exposome, namely the totality of exposures including, among others, diet, lifestyle, and environment, during early life and young adulthood, has changed considerably in the last decades. The categorization of the population into two cohorts with extreme ages (< 50 vs ≥ 70) and the exclusion of those cases with intermediate age favor the identification of substantial generational changes in the exposome of the two cohorts, while those in between may present a gradual change in their exposure. The relevance and impact of this investigation may be multiple, as it suggests research priorities for primary preventive interventions aimed at behavior modifications and secondary prevention in those individuals exposed to risk factors for EOCRC for current and future generations. Also, it might guide the development of health policies for environmental exposures and regulatory policies for agricultural products. Last, the spotlight on the tumor age rather than on the patient is a novel perspective for CRC research and epidemiology for both EOCRC and those diagnosed over the age of 50, and this could be one of the potential explanations for why no differences in tumor biology between EOCRC and LOCRC have been found so far. In conclusion, our findings not only provide exposome traits based on epigenetic fingerprints that could be contributing to the development of CRC, specifically in EOCRC, but also pave the way with a compelling rationale for addressing lifestyle and environmental exposures to mitigate EOCRC risk, highlighting the importance of both personal and policy-level interventions. Methods Study population The discovery phase of our study utilized colon adenocarcinoma (COAD) samples from TCGA. To ensure the integrity of our analysis, patients with Lynch syndrome (TCGA-A6-6781, TCGA-CM-6674, and TCGA-D5-6927) were excluded. Our replication effort encompassed six datasets, focusing on rectal cancer (TCGA-READ and GSE39958 (47)), colon cancer (GSE131013 (45) and GSE42752 (46)), and colorectal cancer studies (GSE101764 (48) and GSE77954 (49)). For the TCGA samples, we implemented exclusion criteria, removing cases annotated with "Item in special subset", "History of unacceptable prior treatment related to a prior/other malignancy", "Case submitted is found to be a recurrence after submission", "Neoadjuvant therapy", "Synchronous malignancy", and "Pathology outside specification". This led to the exclusion of 10 patients from the COAD dataset and 8 from the READ dataset. Additionally, samples preserved in formalin-fixed paraffin-embedded (FFPE) form were excluded to maintain consistency in sample quality. Duplicate samples were identified and removed based on their plate number, further refining our cohorts for analysis. DNA methylation data processing DNA methylation data was in all datasets obtained using the Illumina Infinium HumanMethylation450 BeadChip. TCGA methylation data was acquired from the Genomic Data Commons (GDC) Data Portal through the TCGABiolinks package v2.25.0 (88–90). For the additional datasets, idat files or signal intensity files were downloaded from the Gene Expression Omnibus (GEO) database via the GEOquery package v2.60.0 (91). We processed the tumor sample methylation data with the sesame package v1.19.7 (92–95) in R version 4.3.1. CpG sites showing more than 50% missing data were removed from each dataset. The remaining missing values underwent imputation using the impute package v1.66.0 (96), followed by mean imputation for CpGs with imputed values of 0. DNA methylation beta-values were then converted to M-values using the lumi package v2.44.0 (97–100). Next, we estimated the epigenetic age using the Horvath clock, following the R software tutorial (50), and adjusted the DNA methylation m-values for the estimated age, to account for the influence of age on methylation. Finally, we converted the m-values back to beta-values for further analysis. The final dataset refinement involved excluding CpGs that were not present across all datasets, those associated with cross-reactive probes, and CpGs identified as single nucleotide polymorphisms (101). Exposome-related traits marker selection We conducted a detailed literature review to compile a list of exposome traits potentially influencing the risk of early-onset colon and rectal cancer, ensuring the inclusion of traits with available comprehensive epigenome-wide association studies (EWAS). This process led to the identification of 29 key exposome traits (28–43). To address the challenge of smaller sample sizes in air pollution-related EWAS, we combined CpG sites from multiple studies. The selected exposome traits span a wide range, including lifestyle factors, environmental exposures, and a selection of pesticides. CpG sites were chosen using stringent significance criteria: P<1.2×10 -7 , P<1.0×10 -5 , and false discovery rates (FDR) of <0.01, <0.05, and <0.1. We employed a tiered strategy for marker selection to balance sensitivity and specificity, facilitating a comprehensive assessment of the relationship between exposome exposures, DNA methylation levels, and early-onset colon and rectal cancer. This approach includes a broad range of CpG sites explored using less stringent p-value thresholds (e.g., P<1.0×10 -5 and FDR<0.1), increasing sensitivity to include CpG sites with potential biological relationships. This was instrumental in selecting both highly confident associations and those of potential biological relevance that may not meet the strictest statistical thresholds. In addition, we included CpG sites using more stringent criteria (e.g., P<1.2×10 -7 , FDR<0.05, and FDR<0.01), enhancing specificity and reducing the risk of including false positive CpG sites. This approach allowed for a nuanced exploration of the data, ensuring that both robust and biologically meaningful associated CpG sites were included in our analysis. Exposome-related methylation risk scores in early- vs later-onset patients To assess the impact of these exposome traits on early-onset colon and rectal cancer risk, we computed individual methylation risk scores (MRS) by applying effect sizes from relevant EWAS to methylation levels at each CpG site and summing these across all pertinent CpGs. The obtained scores were normalized to a mean of 0 and a standard deviation of 1, exclusively utilizing the data from EOCRC and LOCRC patients. We examined the correlation between these MRSs and the incidence of early-onset cancer, comparing them to later-onset patients using logistic regression analyses adjusted for sex using the stats v4.1.3. R package. This analysis spanned the discovery dataset TCGA-COAD and extended to six validation cohorts. Models were excluded from the subsequent meta-analysis if they issued fitting warnings or were unable to estimate the 95% confidence interval. The collective data from the validation cohorts were subjected to a meta-analysis utilizing the metafor R package v4.4-0 (102). The obtained p-values were FDR (103) corrected for the number of traits in each marker selection threshold, including 19 traits in GW, 26 in P1E5, 9 in F01, 11 in F005, and 8 in F001, using the p.adjust function in stats v4.1.3. The results were then visually summarized in forest plots, offering a clear depiction of the associations between MRS and the risk of EOCRC relative to later-onset cases. Permutation analysis for CpG selection and patient categorization To evaluate the robustness of our findings, we conducted two permutation analyses, ensuring that the observed associations are not artifacts of the specific CpG selection or patient grouping but reflect underlying biological relationships. The first analysis involved randomly assigning patients to the MRSs. Specifically, we shuffled the early- and later-onset colon cancer patients in the TCGA-COAD dataset 1,000 times, creating alternative configurations of the dataset. For each rearranged dataset, we performed logistic regression analyses, adjusting for sex, to assess the stability of the association between MRSs and early-onset across these permutations. The second analysis targeted the specific assignment of CpGs to their corresponding effect sizes (weights). We temporarily removed the CpGs that were initially selected for each exposome trait to create a pool of potential CpGs. From this pool, we generated 10,000 new sets, each containing an equal number of CpGs as in the original analysis but selected randomly. These randomly chosen CpGs were then used to construct new MRSs for each trait. Subsequently, logistic regression analyses, adjusted for sex, were applied to these permutation-derived MRSs to test the significance of the associations under these randomized conditions. Young vs. older tumors using single-base substitution signature 1 Patients from the TCGA-COAD dataset were selected when presented with comprehensive profiles including DNA methylation data, mutational signatures (https://dcc.icgc.org/releases/PCAWG), and confirmed microsatellite stability. To ensure data integrity, we excluded extreme outlier patients, identified by SBS1 scores exceeding more than three times the standard deviation above the mean SBS1 score. The patients were categorized as early-onset (aged <50), middle-onset (aged 50 to 69 years), and later-onset (aged ≥70 years) cases. Initially, we assessed the differential impact of picloram MRS-GW between the early- and later-onset groups within this subset. Subsequently, we aimed to enhance the specificity of our findings by integrating the concept of tumor age, inferred from the SBS1 score, into our analysis of early- and later-onset categories. In a novel approach, we explored the efficacy of using tumor age—defined by an SBS1 score of 60 as a threshold—as an alternative to chronological age in distinguishing between younger and older tumors. We selected the threshold of an SBS1 score of 60 as this was the first quartile in later-onset patients and the median in middle-onset patients. The third quartile in early-onset patients was an SBS1 score of 67. Pesticide use and early-onset CRC in the United States The results obtained for pesticide MRSs were validated using population data from the U.S. We assessed the association between pesticide use and EOCRC incidence across overlapping counties. Utilizing data from the Pesticide National Synthesis Project, we analyzed county-level pesticide usage between 1992 and 2012 for specific chemicals: acetochlor, 2,4-D, atrazine, dicamba, glyphosate, mesotrione, and picloram. These pesticides were selected based on the availability of complete data. To estimate pesticide use intensity for each county, we divided the total estimated pesticide use by the county's area in square miles. EOCRC incidence rates were obtained from the Surveillance, Epidemiology, and End Results (SEER) database using the SEER*Stat software (version 8.4.1) (104). This comprehensive dataset allowed us to incorporate Research Plus Data spanning from 1975 to 2020 for SEER8 and from 1992 to 2020 for SEER12. We calculated age-adjusted EOCRC incidence rates (for individuals aged 25-49) at the county level annually, normalizing the number of cases per 100,000 population. Counties with no cases for more than 50% of the years were excluded to mitigate the impact of sparse data on our analysis. The pesticide data and EOCRC incidence data were overlapped at county level. To refine our analysis and reduce the influence of extreme values, we removed the top and bottom 5% measurements based on log-transformed pesticide-use intensity. Our analytical model, a linear mixed model, was employed to examine the relationship between pesticide use intensity and EOCRC incidence, adjusting for the years measured and including a random effect to account for variations across counties. We further explored the temporal dynamics of this association by incorporating an interaction term between pesticide use intensity and the measured years. The significance of the interaction term, indicative of changes in the association over time, was assessed using ANOVA for each pesticide under study. Linear mixed models and ANOVA analyses were executed with the lmerTest R package v3.1-3 (105). Statistical analyses were performed in R v4.1.3 (106). Declarations Data availability All methylation and related clinical data used in this study are publicly available via the GDC Data portal (https://portal.gdc.cancer.gov/) for TCGA datasets or via the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/) for GSE131013, GSE42752, GSE39958, GSE101764, and GSE77954. Pesticide use data can be extracted from the National Water-Quality Assessment (NAWQA) project (https://water.usgs.gov/nawqa/pnsp/usage/maps/county-level/). Access to the Research Plus Data from The Surveillance, Epidemiology, and End Results (SEER) cohort can be requested via https://seer.cancer.gov/data/access.html. Code availability The code used for the analysis in this study is publicly available from the following repository: https://github.com/CancerCompBioLab/EOCRCexposome. All the input files needed to replicate our findings and the results obtained during the study are also available from the GitHub repository. Acknowledgments We first and foremost thank the participants and their families included in the used studies. We are grateful to the investigators and data management teams who recruited the participants and to the pathologists who collected the samples. The results shown here are in part based upon data generated by the TCGA Research Network: https://www.cancer.gov/tcga and the Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*Stat Database: Incidence - SEER Plus Research Data, 8 Registries, Nov 2021 Sub (1975-2019) and 12 Registries, Nov 2021 Sub (1992-2019) - Linked To County Attributes based on the November 2021 submission. We thank Javier Carmona, PhD; and the members of the Cancer Computational Biology Group for the helpful discussions. This work was supported by CMS2022-135428, RYC2019-026576-I, PID2020-115097RA-I00, and ISCIII grant FORT23/00034, and Fundacion “la Caixa” (to J.A. Seoane), Juan de la Cierva JDC2022-048829-I (to S.C.E. Maas). Author information Authors and Affiliations Vall d’Hebron Institute of Oncology (VHIO), Barcelona, Catalonia, Spain Silvana C.E. Maas, Iosune Baraibar, Odei Blanco Irazuegui, Josep Tabernero, Elena Elez & Jose A. Seoane Research programme on Biomedical Informatics (GRIB), Hospital del Mar Research Institute, PRBB, Barcelona Odei Blanco Irazuegui Contributions JAS, EE, and SCEM conceived the study design. SCEM and OBI analyzed the data. JAS, EE, JT, SCEM, and IB interpreted the data and results. SCEM, JAS, and IB drafted the paper and SCEM drafted the figures. All authors critically reviewed the paper and the results and approved the final version. Ethics declarations Competing interests IB has received accommodation and travel expenses from Amgen, Merck, Sanofi, and Servier; and personal speaker honoraria from Astra Zeneca. EE has received personal honoraria from Amgen, Bayer, BMS, Boehringer Ingelheim, Cure Teq AG, Hoffman La – Roche, Janssen, Lilly, Medscape, Merck Serono, MSD, Novartis, Organon, Pfizer, Pierre Fabre, Repare Therapeutics Inc., RIN Institute Inc., Sanofi, Seagen International, GmbH, Servier, and Takeda. JT reports personal financial interest in the form of scientific consultancy role for Alentis Therapeutics, AstraZeneca, Aveo Oncology, Boehringer Ingelheim, Cardiff Oncology, CARSgen Therapeutics, Chugai, Daiichi Sankyo, F. Hoffmann-La Roche Ltd, Genentech Inc, hC Bioscience, Ikena Oncology, Immodulon Therapeutics, Inspirna Inc, Lilly, Menarini, Merck Serono, Merus, MSD, Mirati, Neophore, Novartis, Ona Therapeutics, Orion Biotechnology, Peptomyc, Pfizer, Pierre Fabre, Samsung Bioepis, Sanofi, Scandion Oncology, Scorpion Therapeutics, Seattle Genetics, Servier, Sotio Biotech, Taiho, Takeda Oncology and Tolremo Therapeutics. Stocks: Oniria Therapeutics, Alentis Therapeutics, Pangaea Oncology and 1TRIALSP, and also an educational collaboration with Medscape Education, PeerView Institute for Medical Education and Physicians Education Resource (PER). JAS, OBI, and SCEM declare no competing interests. References Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024 Apr 4;74(3):229–63. Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin. 2024 Jan 17;74(1):12–49. Siegel RL, Torre LA, Soerjomataram I, Hayes RB, Bray F, Weber TK, et al. Global patterns and trends in colorectal cancer incidence in young adults. Gut. 2019 Dec;68(12):2179–85. Lui RN, Tsoi KKF, Ho JMW, Lo CM, Chan FCH, Kyaw MH, et al. Global Increasing Incidence of Young-Onset Colorectal Cancer Across 5 Continents: A Joinpoint Regression Analysis of 1,922,167 Cases. Cancer Epidemiol Biomarkers Prev. 2019 Aug;28(8):1275–82. Vuik FE, Nieuwenburg SA, Bardou M, Lansdorp-Vogelaar I, Dinis-Ribeiro M, Bento MJ, et al. Increasing incidence of colorectal cancer in young adults in Europe over the last 25 years. Gut. 2019 Oct;68(10):1820–6. Bailey CE, Hu C-Y, You YN, Bednarski BK, Rodriguez-Bigas MA, Skibber JM, et al. Increasing disparities in the age-related incidences of colon and rectal cancers in the United States, 1975-2010. JAMA Surg. 2015 Jan;150(1):17–22. Akimoto N, Ugai T, Zhong R, Hamada T, Fujiyoshi K, Giannakis M, et al. Rising incidence of early-onset colorectal cancer - a call to action. Nat Rev Clin Oncol. 2021 Apr;18(4):230–43. Zaborowski AM, Murphy B, Creavin B, Rogers AC, Kennelly R, Hanly A, et al. Clinicopathological features and oncological outcomes of patients with young-onset rectal cancer. Br J Surg. 2020 Apr;107(5):606–12. Foppa C, Tamburello S, Maroli A, Carvello M, Poliani L, Laghi L, et al. Early age of onset is an independent predictor for worse disease-free survival in sporadic rectal cancer patients. A comparative analysis of 980 consecutive patients. Eur J Surg Oncol. 2022 Apr;48(4):857–63. McClelland PH-T, Liu T, Ozuner G. Early-Onset Colorectal Cancer in Patients under 50 Years of Age: Demographics, Disease Characteristics, and Survival. Clin Colorectal Cancer. 2022 Jun;21(2):e135–44. Lieu CH, Golemis EA, Serebriiskii IG, Newberg J, Hemmerich A, Connelly C, et al. Comprehensive genomic landscapes in early and later onset colorectal cancer. Clin Cancer Res. 2019 Oct 1;25(19):5852–8. Ugai T, Haruki K, Harrison TA, Cao Y, Qu C, Chan AT, et al. Molecular Characteristics of Early-Onset Colorectal Cancer According to Detailed Anatomical Locations: Comparison With Later-Onset Cases. Am J Gastroenterol. 2023 Apr 1;118(4):712–26. Berg M, Agesen TH, Thiis-Evensen E, INFAC-study group, Merok MA, Teixeira MR, et al. Distinct high resolution genome profiles of early onset and late onset colorectal cancer integrated with gene expression data identify candidate susceptibility loci. Mol Cancer. 2010 May 6;9:100. Willauer AN, Liu Y, Pereira AAL, Lam M, Morris JS, Raghav KPS, et al. Clinical and molecular characterization of early-onset colorectal cancer. Cancer. 2019 Jun 15;125(12):2002–10. Perea J, Rueda D, Canal A, Rodríguez Y, Álvaro E, Osorio I, et al. Age at onset should be a major criterion for subclassification of colorectal cancer. J Mol Diagn. 2014 Jan;16(1):116–26. Antelo M, Balaguer F, Shia J, Shen Y, Hur K, Moreira L, et al. A high degree of LINE-1 hypomethylation is a unique feature of early-onset colorectal cancer. PLoS ONE. 2012 Sep 25;7(9):e45357. Baba Y, Huttenhower C, Nosho K, Tanaka N, Shima K, Hazra A, et al. Epigenomic diversity of colorectal cancer indicated by LINE-1 methylation in a database of 869 tumors. Mol Cancer. 2010 May 27;9:125. Akimoto N, Zhao M, Ugai T, Zhong R, Lau MC, Fujiyoshi K, et al. Tumor Long Interspersed Nucleotide Element-1 (LINE-1) Hypomethylation in Relation to Age of Colorectal Cancer Diagnosis and Prognosis. Cancers (Basel). 2021 Apr 22;13(9). Gogna P, O’Sullivan DE, King WD. The effect of inflammation-related lifestyle exposures and interactions with gene variants on long interspersed nuclear element-1 DNA methylation. Epigenomics. 2018 Jun 11;10(6):785–96. Martin EM, Fry RC. Environmental influences on the epigenome: Exposure-associated DNA methylation in human populations. Annu Rev Public Health. 2018 Apr 1;39:309–33. Cho YH, Jang Y, Woo HD, Kim YJ, Kim SY, Christensen S, et al. LINE-1 hypomethylation is associated with radiation-induced genomic instability in industrial radiographers. Environ Mol Mutagen. 2019 Mar;60(2):174–84. Mork ME, You YN, Ying J, Bannon SA, Lynch PM, Rodriguez-Bigas MA, et al. High prevalence of hereditary cancer syndromes in adolescents and young adults with colorectal cancer. J Clin Oncol. 2015 Nov 1;33(31):3544–9. Stoffel EM, Koeppe E, Everett J, Ulintz P, Kiel M, Osborne J, et al. Germline genetic features of young individuals with colorectal cancer. Gastroenterology. 2018 Mar;154(4):897-905.e1. Syed AR, Thakkar P, Horne ZD, Abdul-Baki H, Kochhar G, Farah K, et al. Old vs new: Risk factors predicting early onset colorectal cancer. World J Gastrointest Oncol. 2019 Nov 15;11(11):1011–20. Connell LC, Mota JM, Braghiroli MI, Hoff PM. The rising incidence of younger patients with colorectal cancer: questions about screening, biology, and treatment. Curr Treat Options Oncol. 2017 Apr;18(4):23. Pleguezuelos-Manzano C, Puschhof J, Rosendahl Huber A, van Hoeck A, Wood HM, Nomburg J, et al. Mutational signature in colorectal cancer caused by genotoxic pks+ E. coli. Nature. 2020 Apr;580(7802):269–73. Rosendahl Huber A, Pleguezuelos-Manzano C, Puschhof J, Ubels J, Boot C, Saftien A, et al. Improved detection of colibactin-induced mutations by genotoxic E. coli in organoids and colorectal cancer. Cancer Cell. 2024 Mar 11;42(3):487-496.e6. Ma J, Rebholz CM, Braun KVE, Reynolds LM, Aslibekyan S, Xia R, et al. Whole Blood DNA Methylation Signatures of Diet Are Associated With Cardiovascular Disease Risk Factors and All-Cause Mortality. Circ Genom Precis Med. 2020 Aug;13(4):e002766. Liu C, Marioni RE, Hedman ÅK, Pfeiffer L, Tsai PC, Reynolds LM, et al. A DNA methylation biomarker of alcohol consumption. Mol Psychiatry. 2018 Feb;23(2):422–33. Madden RA, McCartney DL, Walker RM, Hillary RF, Bermingham ML, Rawlik K, et al. Birth weight associations with DNA methylation differences in an adult population. Epigenetics. 2021;16(7):783–96. Do WL, Sun D, Meeks K, Dugué P-A, Demerath E, Guan W, et al. Epigenome-wide meta-analysis of BMI in nine cohorts: Examining the utility of epigenetically predicted BMI. Am J Hum Genet. 2023 Feb 2;110(2):273–83. Karabegović I, Portilla-Fernandez E, Li Y, Ma J, Maas SCE, Sun D, et al. Epigenome-wide association meta-analysis of DNA methylation with coffee and tea consumption. Nat Commun. 2021 May 14;12(1):2830. van Dongen J, Bonder MJ, Dekkers KF, Nivard MG, van Iterson M, Willemsen G, et al. DNA methylation signatures of educational attainment. NPJ Sci Learn. 2018 Mar 23;3:7. Koh I-U, Choi N-H, Lee K, Yu H-Y, Yun JH, Kong J-H, et al. Obesity susceptible novel DNA methylation marker on regulatory region of inflammation gene: results from the Korea Epigenome Study (KES). BMJ Open Diabetes Res Care. 2020 Aug;8(1). Joehanes R, Just AC, Marioni RE, Pilling LC, Reynolds LM, Mandaviya PR, et al. Epigenetic signatures of cigarette smoking. Circ Cardiovasc Genet. 2016 Oct;9(5):436–47. Maas SCE, Vidaki A, Wilson R, Teumer A, Liu F, van Meurs JBJ, et al. Validated inference of smoking habits from blood with a finite DNA methylation marker set. Eur J Epidemiol. 2019 Nov;34(11):1055–74. de F C Lichtenfels AJ, van der Plaat DA, de Jong K, van Diemen CC, Postma DS, Nedeljkovic I, et al. Long-term Air Pollution Exposure, Genome-wide DNA Methylation and Lung Function in the LifeLines Cohort Study. Environ Health Perspect. 2018 Feb 6;126(2):027004. Plusquin M, Guida F, Polidoro S, Vermeulen R, Raaschou-Nielsen O, Campanella G, et al. DNA methylation and exposure to ambient air pollution in two prospective cohorts. Environ Int. 2017 Nov;108:127–36. Sayols-Baixeras S, Fernández-Sanlés A, Prats-Uribe A, Subirana I, Plusquin M, Künzli N, et al. Association between long-term air pollution exposure and DNA methylation: The REGICOR study. Environ Res. 2019 Sep;176:108550. Gondalia R, Baldassari A, Holliday KM, Justice AE, Méndez-Giráldez R, Stewart JD, et al. Methylome-wide association study provides evidence of particulate matter air pollution-associated DNA methylation. Environ Int. 2019 Nov;132:104723. Panni T, Mehta AJ, Schwartz JD, Baccarelli AA, Just AC, Wolf K, et al. Genome-Wide Analysis of DNA Methylation and Fine Particulate Matter Air Pollution in Three Study Populations: KORA F3, KORA F4, and the Normative Aging Study. Environ Health Perspect. 2016 Jul;124(7):983–90. Curtis SW, Cobb DO, Kilaru V, Terrell ML, Marder ME, Barr DB, et al. Genome-wide DNA methylation differences and polychlorinated biphenyl (PCB) exposure in a US population. Epigenetics. 2021 Mar;16(3):338–52. Hoang TT, Qi C, Paul KC, Lee M, White JD, Richards M, et al. Epigenome-Wide DNA Methylation and Pesticide Use in the Agricultural Lung Health Study. Environ Health Perspect. 2021 Sep 13;129(9):97008. Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature. 2012 Jul 18;487(7407):330–7. Díez-Villanueva A, Sanz-Pamplona R, Carreras-Torres R, Moratalla-Navarro F, Alonso MH, Paré-Brunet L, et al. DNA methylation events in transcription factors and gene expression changes in colon cancer. Epigenomics. 2020 Sep 22;12(18):1593–610. Naumov VA, Generozov EV, Zaharjevskaya NB, Matushkina DS, Larin AK, Chernyshov SV, et al. Genome-scale analysis of DNA methylation in colorectal cancer using Infinium HumanMethylation450 BeadChips. Epigenetics. 2013 Sep;8(9):921–34. Ha YJ, Kim CW, Roh SA, Cho DH, Park JL, Kim SY, et al. Epigenetic regulation of KLHL34 predictive of pathologic response to preoperative chemoradiation therapy in rectal cancer patients. Int J Radiat Oncol Biol Phys. 2015 Mar 1;91(3):650–8. Barrow TM, Klett H, Toth R, Böhm J, Gigic B, Habermann N, et al. Smoking is associated with hypermethylation of the APC 1A promoter in colorectal cancer: the ColoCare Study. J Pathol. 2017 Nov;243(3):366–75. Qu X, Sandmann T, Frierson H, Fu L, Fuentes E, Walter K, et al. Integrated genomic analysis of colorectal cancer progression reveals activation of EGFR through demethylation of the EREG promoter. Oncogene. 2016 Dec 15;35(50):6403–15. Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14(10):R115. Ramai D, Barakat M, Dhaliwal A, Dhindsa B, Chandan S, Ofosu A, et al. Gender and racial disparities in colorectal cancer incidence and mortality: a national cancer registry study. Int J Colorectal Dis. 2021 Aug;36(8):1801–4. Zhong Y, Zhu Y, Li Q, Wang F, Ge X, Zhou G, et al. Association between Mediterranean diet adherence and colorectal cancer: a dose-response meta-analysis. Am J Clin Nutr. 2020 Jun 1;111(6):1214–25. Doubeni CA, Laiyemo AO, Major JM, Schootman M, Lian M, Park Y, et al. Socioeconomic status and the risk of colorectal cancer: an analysis of more than a half million adults in the National Institutes of Health-AARP Diet and Health Study. Cancer. 2012 Jul 15;118(14):3636–44. Botteri E, Borroni E, Sloan EK, Bagnardi V, Bosetti C, Peveri G, et al. Smoking and Colorectal Cancer Risk, Overall and by Molecular Subtypes: A Meta-Analysis. Am J Gastroenterol. 2020 Dec;115(12):1940–9. Mandic M, Safizadeh F, Niedermaier T, Hoffmeister M, Brenner H. Association of overweight, obesity, and recent weight loss with colorectal cancer risk. JAMA Netw Open. 2023 Apr 3;6(4):e239556. Alexandrov LB, Jones PH, Wedge DC, Sale JE, Campbell PJ, Nik-Zainal S, et al. Clock-like mutational processes in human somatic cells. Nat Genet. 2015 Dec;47(12):1402–7. Alexandrov LB, Kim J, Haradhvala NJ, Huang MN, Tian Ng AW, Wu Y, et al. The repertoire of mutational signatures in human cancer. Nature. 2020 Feb 5;578(7793):94–101. Sia EA, Kokoska RJ, Dominska M, Greenwell P, Petes TD. Microsatellite instability in yeast: dependence on repeat unit size and DNA mismatch repair genes. Mol Cell Biol. 1997 May;17(5):2851–8. Boland CR, Goel A. Microsatellite instability in colorectal cancer. Gastroenterology. 2010 Jun;138(6):2073-2087.e3. Arnold M, Abnet CC, Neale RE, Vignat J, Giovannucci EL, McGlynn KA, et al. Global burden of 5 major types of gastrointestinal cancer. Gastroenterology. 2020 Jul;159(1):335-349.e15. Huyghe JR, Bien SA, Harrison TA, Kang HM, Chen S, Schmit SL, et al. Discovery of common and rare genetic risk variants for colorectal cancer. Nat Genet. 2019 Jan;51(1):76–87. Steele SR, Park GE, Johnson EK, Martin MJ, Stojadinovic A, Maykel JA, et al. The impact of age on colorectal cancer incidence, treatment, and outcomes in an equal-access health care system. Dis Colon Rectum. 2014 Mar;57(3):303–10. Chen N, Liang H, Huang T, Huang N. Exposome approach for identifying modifiable factors for the prevention of colorectal cancer. Sci Rep. 2022 Dec 14;12(1):21615. Carr PR, Weigl K, Edelmann D, Jansen L, Chang-Claude J, Brenner H, et al. Estimation of Absolute Risk of Colorectal Cancer Based on Healthy Lifestyle, Genetic Risk, and Colonoscopy Status in a Population-Based Study. Gastroenterology. 2020 Jul;159(1):129-138.e9. Huang P, Liu Y. A reasonable diet promotes balance of intestinal microbiota: prevention of precolorectal cancer. Biomed Res Int. 2019 Jul 25;2019:3405278. Wilkins T, McMechan D, Talukder A. Colorectal cancer screening and prevention. Am Fam Physician. 2018 May 15;97(10):658–65. Brooke HL, Talbäck M, Martling A, Feychting M, Ljung R. Socioeconomic position and incidence of colorectal cancer in the Swedish population. Cancer Epidemiol. 2016 Feb;40:188–95. Egeberg R, Halkjaer J, Rottmann N, Hansen L, Holten I. Social inequality and incidence of and survival from cancers of the colon and rectum in a population-based study in Denmark, 1994-2003. Eur J Cancer. 2008 Sep;44(14):1978–88. Sarma EA, Kawachi I, Poole EM, Tworoger SS, Giovannucci EL, Fuchs CS, et al. Social integration and survival after diagnosis of colorectal cancer. Cancer. 2018 Feb 15;124(4):833–40. García-Pérez J, Fernández de Larrea-Baz N, Lope V, Molina AJ, O’Callaghan-Gordo C, Alonso MH, et al. Residential proximity to industrial pollution sources and colorectal cancer risk: A multicase-control study (MCC-Spain). Environ Int. 2020 Nov;144:106055. Jenwitheesuk K, Peansukwech U, Jenwitheesuk K. Accumulated ambient air pollution and colon cancer incidence in Thailand. Sci Rep. 2020 Oct 20;10(1):17765. Pritchett N, Spangler EC, Gray GM, Livinski AA, Sampson JN, Dawsey SM, et al. Exposure to Outdoor Particulate Matter Air Pollution and Risk of Gastrointestinal Cancers in Adults: A Systematic Review and Meta-Analysis of Epidemiologic Evidence. Environ Health Perspect. 2022 Mar 2;130(3):36001. López-Abente G, García-Pérez J, Fernández-Navarro P, Boldo E, Ramis R. Colorectal cancer mortality and industrial pollution in Spain. BMC Public Health. 2012 Aug 1;12:589. Wei Y, Danesh Yazdi M, Ma T, Castro E, Liu CS, Qiu X, et al. Additive effects of 10-year exposures to PM2.5 and NO2 and primary cancer incidence in American older adults. Environ Epidemiol. 2023 Aug 1;7(4):e265. Feng J, Cavallero S, Hsiai T, Li R. Impact of air pollution on intestinal redox lipidome and microbiome. Free Radic Biol Med. 2020 May 1;151:99–110. Jin Y, Lu L, Tu W, Luo T, Fu Z. Impacts of polystyrene microplastic on the gut barrier, microbiota and metabolism of mice. Sci Total Environ. 2019 Feb 1;649:308–17. Xia T, Guo J, Zhang B, Song C, Zhao Q, Cui B, et al. Bisphenol A Promotes the Progression of Colon Cancer Through Dual-Targeting of NADPH Oxidase and Mitochondrial Electron-Transport Chain to Produce ROS and Activating HIF-1α/VEGF/PI3K/AKT Axis. Front Endocrinol (Lausanne). 2022 Jul 4;13:933051. United States Environmental Protection Agency. US EPA - Pesticides - Fact Sheet for Picloram. 1995 Aug [cited 2024 May 13]; Available from: https://www3.epa.gov/pesticides/chem_search/reg_actions/reregistration/fs_PC-005101_1-Aug-95.pdf Davoren MJ, Schiestl RH. Glyphosate-based herbicides and cancer risk: a post-IARC decision review of potential mechanisms, policy and avenues of research. Carcinogenesis. 2018 Oct 8;39(10):1207–15. Sass JB, Colangelo A. European Union bans atrazine, while the United States negotiates continued use. Int J Occup Environ Health. 2006 Sep;12(3):260–7. Turner MC, Cogliano V, Guyton K, Madia F, Straif K, Ward EM, et al. Research Recommendations for Selected IARC-Classified Agents: Impact and Lessons Learned. Environ Health Perspect. 2023 Oct 30;131(10):105001. Berrington de González A, Masten SA, Bhatti P, Fortner RT, Peters S, Santonen T, et al. Advisory Group recommendations on priorities for the IARC Monographs. Lancet Oncol. 2024 May;25(5):546–8. Remigio RV, Andreotti G, Sandler DP, Erickson PA, Koutros S, Albert PS, et al. An Updated Evaluation of Atrazine-Cancer Incidence Associations among Pesticide Applicators in the Agricultural Health Study Cohort. Environ Health Perspect. 2024 Feb 21;132(2):27010. O’Sullivan DE, Sutherland RL, Town S, Chow K, Fan J, Forbes N, et al. Risk Factors for Early-Onset Colorectal Cancer: A Systematic Review and Meta-analysis. Clin Gastroenterol Hepatol. 2022 Jun;20(6):1229-1240.e5. Cercek A, Chatila WK, Yaeger R, Walch H, Fernandes GDS, Krishnan A, et al. A Comprehensive Comparison of Early-Onset and Average-Onset Colorectal Cancers. J Natl Cancer Inst. 2021 Nov 29;113(12):1683–92. Song N, Shin A, Park JW, Kim J, Oh JH. Common risk variants for colorectal cancer: an evaluation of associations with age at cancer onset. Sci Rep. 2017 Jan 13;7:40644. Chen FW, Sundaram V, Chew TA, Ladabaum U. Advanced-Stage Colorectal Cancer in Persons Younger Than 50 Years Not Associated With Longer Duration of Symptoms or Time to Diagnosis. Clin Gastroenterol Hepatol. 2017 May;15(5):728-737.e3. Silva TC, Colaprico A, Olsen C, D’Angelo F, Bontempi G, Ceccarelli M, et al. TCGA Workflow: Analyze cancer genomics and epigenomics data using Bioconductor packages. [version 2; peer review: 1 approved, 2 approved with reservations]. F1000Res. 2016 Jun 29;5:1542. Colaprico A, Silva TC, Olsen C, Garofano L, Cava C, Garolini D, et al. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res. 2016 May 5;44(8):e71. Mounir M, Lucchetta M, Silva TC, Olsen C, Bontempi G, Chen X, et al. New functionalities in the TCGAbiolinks package for the study and integration of cancer data from GDC and GTEx. PLoS Comput Biol. 2019 Mar 5;15(3):e1006701. Davis S, Meltzer PS. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics. 2007 Jul 15;23(14):1846–7. Ding W, Kaur D, Horvath S, Zhou W. Comparative epigenome analysis using Infinium DNA methylation BeadChips. Brief Bioinformatics. 2023 Jan 19;24(1). Zhou W, Hinoue T, Barnes B, Mitchell O, Iqbal W, Lee SM, et al. DNA methylation dynamics and dysregulation delineated by high-throughput profiling in the mouse. Cell Genomics. 2022 Jul 13;2(7). Zhou W, Triche TJ, Laird PW, Shen H. SeSAMe: reducing artifactual detection of DNA methylation by Infinium BeadChips in genomic deletions. Nucleic Acids Res. 2018 Nov 16;46(20):e123. Triche TJ, Weisenberger DJ, Van Den Berg D, Laird PW, Siegmund KD. Low-level processing of Illumina Infinium DNA Methylation BeadArrays. Nucleic Acids Res. 2013 Apr;41(7):e90. Hastie T, Tibshirani R, Narasimhan B, Chu G. Impute: Imputation for microarray data. 2016 Jan 1;17:520–5. Du P, Kibbe WA, Lin SM. lumi: a pipeline for processing Illumina microarray. Bioinformatics. 2008 Jul 1;24(13):1547–8. Du P, Zhang X, Huang C-C, Jafari N, Kibbe WA, Hou L, et al. Comparison of Beta-value and M-value methods for quantifying methylation levels by microarray analysis. BMC Bioinformatics. 2010 Nov 30;11:587. Lin SM, Du P, Huber W, Kibbe WA. Model-based variance-stabilizing transformation for Illumina microarray data. Nucleic Acids Res. 2008 Feb;36(2):e11. Du P, Kibbe WA, Lin SM. nuID: a universal naming scheme of oligonucleotides for illumina, affymetrix, and other microarrays. Biol Direct. 2007 May 31;2:16. Chen Y, Lemire M, Choufani S, Butcher DT, Grafodatskaya D, Zanke BW, et al. Discovery of cross-reactive probes and polymorphic CpGs in the Illumina Infinium HumanMethylation450 microarray. Epigenetics. 2013 Feb;8(2):203–9. Viechtbauer W. Conducting Meta-Analyses in R with the metafor Package. J Stat Softw. 2010;36(3). Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society: Series B (Methodological). 1995 Jan;57(1):289–300. Surveillance Research Program, National Cancer Institute. Surveillance Research Program, National Cancer Institute SEER*Stat software (seer.cancer.gov/seerstat) version 8.4.1. Kuznetsova A, Brockhoff PB, Christensen RHB. lmertest package: tests in linear mixed effects models. J Stat Softw. 2017;82(13):1–26. Team RRC. R: A language and environment for statistical computing. 2013; Additional Declarations Yes there is potential Competing Interest. IB has received accommodation and travel expenses from Amgen, Merck, Sanofi, and Servier; and personal speaker honoraria from Astra Zeneca. EE has received personal honoraria from Amgen, Bayer, BMS, Boehringer Ingelheim, Cure Teq AG, Hoffman La – Roche, Janssen, Lilly, Medscape, Merck Serono, MSD, Novartis, Organon, Pfizer, Pierre Fabre, Repare Therapeutics Inc., RIN Institute Inc., Sanofi, Seagen International, GmbH, Servier, and Takeda. JT reports personal financial interest in the form of scientific consultancy role for Alentis Therapeutics, AstraZeneca, Aveo Oncology, Boehringer Ingelheim, Cardiff Oncology, CARSgen Therapeutics, Chugai, Daiichi Sankyo, F. Hoffmann-La Roche Ltd, Genentech Inc, hC Bioscience, Ikena Oncology, Immodulon Therapeutics, Inspirna Inc, Lilly, Menarini, Merck Serono, Merus, MSD, Mirati, Neophore, Novartis, Ona Therapeutics, Orion Biotechnology, Peptomyc, Pfizer, Pierre Fabre, Samsung Bioepis, Sanofi, Scandion Oncology, Scorpion Therapeutics, Seattle Genetics, Servier, Sotio Biotech, Taiho, Takeda Oncology and Tolremo Therapeutics. Stocks: Oniria Therapeutics, Alentis Therapeutics, Pangaea Oncology and 1TRIALSP, and also an educational collaboration with Medscape Education, PeerView Institute for Medical Education and Physicians Education Resource (PER). JAS, OBI, and SCEM declare no competing interests. Supplementary Files FigS1heatmapExpl.pdf Supplementary Figure S1: Directionality of the methylation risk scores FigS2obesity.pdf Supplementary Figure S2: Later-onset colon cancer patients are more likely to be obese than early-onset patients. FigS3heatmapallmeta.pdf Supplementary Figure S3: Exposome-related methylation risk scores show differences between early-onset and later-onset colon and rectal cancer patients FigS4Permutationposcontr.pdf Supplementary Figure S4: Permutation results for CpG sites and patient categorization FigS5SBS1MSIstat.pdf Supplementary Figure S5: SBS1-score is correlated with chronological age. SupplementarytablesEpigeneticFingerprintsLinkEarlyOnsetColonandRectalCancertoPesticideExposure04062024.xlsx Supplementary Tables S1–S5 This workbook contains 5 Supplementary tables: Table S1 shows an overview of the included CpGs per exposome trait per marker selection threshold; Table S2 provides the CpGs and their respective weights per trait; Table S3 presents the results obtained in the discovery phase (TCGA-COAD) for all thresholds; Table S4 presents the results obtained in the replication meta-analysis CRC; and Table S5 presents the results obtained in TCGA-READ and the meta-analysis of colon and rectal cancer datasets. Cite Share Download PDF Status: Published Journal Publication published 21 Apr, 2026 Read the published version in Nature Medicine → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4528579","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Analysis","associatedPublications":[],"authors":[{"id":312992698,"identity":"75c1910f-0355-4b01-b3d3-9a1e53f5a9ae","order_by":0,"name":"Silvana Maas","email":"","orcid":"","institution":"Vall d'Hebron Institute of Oncology (VHIO)","correspondingAuthor":false,"prefix":"","firstName":"Silvana","middleName":"","lastName":"Maas","suffix":""},{"id":312992699,"identity":"460a0323-9947-4466-be6a-3f6e5475089b","order_by":1,"name":"Iosune Baraibar","email":"","orcid":"","institution":"Vall Hebron University Hospital and Vall Hebron Institute of Oncology (VHIO)","correspondingAuthor":false,"prefix":"","firstName":"Iosune","middleName":"","lastName":"Baraibar","suffix":""},{"id":312992700,"identity":"800eb92f-bf22-4c0b-bf5d-1a081b877f51","order_by":2,"name":"Odei Blanco-Irazuegui","email":"","orcid":"https://orcid.org/0000-0002-9989-0159","institution":"Hospital del Mar Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Odei","middleName":"","lastName":"Blanco-Irazuegui","suffix":""},{"id":312992701,"identity":"4ba8a596-611a-431a-ac58-d03f101d6c10","order_by":3,"name":"Josep Tabernero","email":"","orcid":"https://orcid.org/0000-0002-2495-8139","institution":"Vall d’Hebron University Hospital and Institute of Oncology (VHIO)","correspondingAuthor":false,"prefix":"","firstName":"Josep","middleName":"","lastName":"Tabernero","suffix":""},{"id":312992702,"identity":"a4638925-acb4-43e2-8bbc-9ac037eb7cff","order_by":4,"name":"Elena Elez","email":"","orcid":"https://orcid.org/0000-0002-4653-6324","institution":"Vall d´Hebron Institute of Oncology (VHIO)","correspondingAuthor":false,"prefix":"","firstName":"Elena","middleName":"","lastName":"Elez","suffix":""},{"id":312992697,"identity":"8a75ec88-de97-4928-bcc5-614364e4b848","order_by":5,"name":"Jose Seoane","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3ElEQVRIiWNgGAWjYBACxgYk9gPGBgnStDAbQLUgC+IHbBJQtfi1MLe3P3zw44+NHP/s5mPVvDss5Bikm48/wOuwnjPGhr1tacYSd46l3eY9I2HMIHMsEa8tjDNy2CR4Gw4nNtzIMbs5s00isUEix5CAlvTnP//8+Z84/0b+t0KIlvyPBLQkmDHzsB1I3HAjh43hI8QW/N4H+UVati3Z2PAG0D9ALcZsEmmGM/BpMQSG2Mc3f+zk5G4kP/yQ2FYnxy+R/OADXi0YbmDDpxwE5AkpGAWjYBSMglHAAAA/6ExbcDRYiAAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-3856-9177","institution":"Vall d'Hebron Institute of Oncology (VHIO)","correspondingAuthor":true,"prefix":"","firstName":"Jose","middleName":"","lastName":"Seoane","suffix":""}],"badges":[],"createdAt":"2024-06-04 13:56:04","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4528579/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4528579/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41591-026-04342-5","type":"published","date":"2026-04-21T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":58223297,"identity":"fff24145-be3a-48fd-a049-5357f0aab866","added_by":"auto","created_at":"2024-06-12 17:17:26","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":55482,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eExposome-related methylation risk scores show differences between early- and later-onset colon and rectal cancer patients\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSummary overview of the association between exposome-related methylation risk scores (MRS) comparing early-onset (age \u0026lt;50) to later-onset cases (age ≥70), where the later-onset category is set as reference. The discovery phase was conducted in TCGA-COAD and includes 31 early-onset and 100 later-onset colon cancer cases, the obtained results are presented in the left panel. The replication phase comprises the meta-analysis of the replication cohorts (GSE131013, GSE42752, TCGA-READ, GSE39958, GSE101764, and GSE77954) which includes 48 patients with early-onset and 151 with later-onset colon and rectal cancer, the obtained results are presented in the right panel. The marker selection thresholds indicate the significance threshold in the original EWAS used for CpG selection, namely P\u0026lt;1.2×10\u003csup\u003e-7 \u003c/sup\u003e(GW), P\u0026lt;1.0×10\u003csup\u003e-5 \u003c/sup\u003e(P1E5), and false discovery rates of \u0026lt;0.1 (F01), \u0026lt;0.05 (F005), and \u0026lt;00.1 (F001). The presence of CpGs across the five marker selection thresholds varies, with unavailable data indicated in white as N.A., while non-significant findings are represented in light grey. 2,4-D; 2,4-Dichlorophenoxyacetic acid, AHEI; Alternative Healthy Eating Index, BMI; body mass index, DDT; Dichlorodiphenyltrichloroethane, MDS; Mediterranean Diet Score, NO2; nitrogen dioxide, PCB; polychlorinated biphenyls, PM10, PM2.5, and PM2.5-10; particulate matter \u0026lt;10, \u0026lt;2.5, and between 2.5 and 10 micrometers (µm) in diameter, smoking-Maas; CpGs from smoking inference model.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/9ed0752db08995cc07fcaa2a.png"},{"id":58222965,"identity":"6fa075e9-f2cc-4774-b4f8-2948885de3b9","added_by":"auto","created_at":"2024-06-12 17:09:26","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":129724,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLifestyle factors in early-onset vs. later-onset colon and rectal cancer tumors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLifestyle-related methylation risk score (MRS) differences were identified between patients with early-onset and later-onset colon and rectal cancer, including \u003cstrong\u003eA.\u003c/strong\u003e non-Mediterranean dietary patterns (MRS-GW [marker threshold P\u0026lt;1.2×10\u003csup\u003e-7\u003c/sup\u003e]), \u003cstrong\u003eB.\u003c/strong\u003e lower education levels, \u003cstrong\u003eC.\u003c/strong\u003e higher smoking exposure, and \u003cstrong\u003eD\u003c/strong\u003e. lower obesity rates (MRS-P1E5 [marker threshold P\u0026lt;1.0×10\u003csup\u003e-5\u003c/sup\u003e]) in early-onset patients. The heatmaps (left) display epigenetic age-adjusted DNA methylation beta-values in TCGA-COAD patients sorted based on the MRS, the color bar above the heatmap reflects the CpG's association direction from the original EWAS; red for positive effect sizes and blue for negative effect sizes. The boxplots (middle) show the distribution of the MRS in early and later-onset patients in TCGA-COAD. As the regressions are corrected for sex, we show the distribution for females (orange) and males (purple) separately. The forest plots (right) show the result obtained in each replication dataset and their meta-analysis (Overall: replication) and combined with TCGA-COAD (Overall: Discovery and Replication), including colon cancer patients (blue), rectal cancer patients (purple), and colorectal patients (red). COAD; colon tumor samples in TCGA-COAD, READ; rectal tumor samples in TCGA-READ, MRS; methylation risk score, OR; Odds Ratio, CI; Confidence interval.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/910af58bb3a6aaa68fc20a63.png"},{"id":58222968,"identity":"e2f98c8b-ef11-4c44-a84d-25daeb407eb7","added_by":"auto","created_at":"2024-06-12 17:09:26","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":66842,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA higher picloram methylation risk score is associated with early-onset colon and rectal cancer.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe methylation risk scores for picloram for all four available marker selection thresholds show a consistently significant association with early-onset colon and rectal cancer. The figure shows the data distribution and obtained results for the methylation risk score employing the genome-wide marker selection threshold (MRS-GW). \u003cstrong\u003eA.\u003c/strong\u003e The heatmap displays epigenetic age-adjusted DNA methylation beta-values in TCGA-COAD patients sorted on the MRS. The color bar above the heatmap reflects the CpG's association direction from the original EWAS: red for positive effect sizes and blue for negative effect sizes. \u003cstrong\u003eB.\u003c/strong\u003e The boxplot shows the distribution of the picloram MRS-GW in early and later onset patients in TCGA-COAD for females (orange) and males (purple). \u003cstrong\u003eC.\u003c/strong\u003e The forest plot shows the result obtained in each replication dataset and their meta-analysis (Overall: Replication) and combined with TCGA-COAD (Overall: Discovery and Replication), including colon cancer patients (blue), rectal cancer patients (purple), and colorectal patients (orange). \u003cstrong\u003eD.\u003c/strong\u003e The violin plot shows the results from the permutation tests conducted to test the stability of the CpGs (left) and the patient categorization (right). The CpG sites included in our model ranked 13\u003csup\u003eth\u003c/sup\u003e in significance out of the 10,000 permutations while the patient categorization ranked 2\u003csup\u003end\u003c/sup\u003e out of the 1,000 permutations. \u003cstrong\u003eE. \u003c/strong\u003eSingle-base substitution signature 1 (SBS1) score distribution in patient categorization based on age at diagnosis, including early-onset (aged \u0026lt;50), middle-onset (aged 50 to 69 years), and later-onset (aged ≥70 years) cases. \u003cstrong\u003eF.\u003c/strong\u003e\u0026nbsp; Patients were categorized based on the acquired mutations belonging to SBS1, where patients with an SBS1 score \u0026lt;60 were classified as young tumors and patients with a score ≥60 as old tumors. The boxplots present the picloram-MRS distribution for the SBS1-based categorization and \u003cstrong\u003eG. \u003c/strong\u003ethe chronological age distribution within the young and older tumors based on the SBS1 categorization. COAD; colon tumor samples in TCGA-COAD, READ; rectal tumor samples in TCGA-READ, MRS; methylation risk score, OR; Odds Ratio, CI; Confidence interval.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/661db9893e4289fc0db7f916.png"},{"id":58222967,"identity":"644cc5ee-be66-4d96-9884-613d140eb390","added_by":"auto","created_at":"2024-06-12 17:09:26","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":43040,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePesticide use intensity is associated with early-onset colorectal cancer incidence.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe use of seven pesticides on county level was obtained between 1992 and 2012 and overlapped with counties with available EOCRC incidence data. \u003cstrong\u003eA. \u003c/strong\u003eThe tile plots show per pesticide the measured years and the number of included counties in SEER8 (blue) and SEER12 (orange). \u003cstrong\u003eB. \u003c/strong\u003eThe maps of Iowa depict the average EOCRC incidence rates (IR CRC) (top) and the log transformation from the average picloram use intensity (bottom) between 1992 and 2012 for each available county. \u003cstrong\u003eC. \u003c/strong\u003eThe bar plot shows the –log\u003csub\u003e10\u003c/sub\u003e p-value of the association between pesticide use intensity and EOCRC incidence of each pesticide in SEER8 (blue) and SEER12 (orange), ordered by results obtained in SEER8. The red dotted line depicts –log\u003csub\u003e10\u003c/sub\u003e(P = 0.05).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/908cbaa83d771302cb7e0850.png"},{"id":107603974,"identity":"eafb6265-c9c5-46a6-bdfc-8e4fbc148f84","added_by":"auto","created_at":"2026-04-23 07:15:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":799856,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/6470b0eb-22f5-4af6-93f0-4700101df1e0.pdf"},{"id":58222974,"identity":"d272a9b0-6ed8-467c-b7d1-17be2293a425","added_by":"auto","created_at":"2024-06-12 17:09:26","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":134134,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Figure S1: Directionality of the methylation risk scores\u003c/p\u003e","description":"","filename":"FigS1heatmapExpl.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/7143b329c833951a2f8b2f26.pdf"},{"id":58222970,"identity":"58346c3a-1c52-4943-8aad-29db39ffde71","added_by":"auto","created_at":"2024-06-12 17:09:26","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":102391,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Figure S2: Later-onset colon cancer patients are more likely to be obese than early-onset patients.\u003c/p\u003e","description":"","filename":"FigS2obesity.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/415e87a7f7a93f8635058530.pdf"},{"id":58222969,"identity":"51be4400-2815-4ad7-af99-06ef23e8af9c","added_by":"auto","created_at":"2024-06-12 17:09:26","extension":"pdf","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":8056,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Figure S3: Exposome-related methylation risk scores show differences between early-onset and later-onset colon and rectal cancer patients\u003c/p\u003e","description":"","filename":"FigS3heatmapallmeta.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/48649d8682b70a53db01c8cf.pdf"},{"id":58223299,"identity":"75c941f2-81a0-404b-98f4-079c2b7ad280","added_by":"auto","created_at":"2024-06-12 17:17:26","extension":"pdf","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":1093296,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Figure S4: Permutation results for CpG sites and patient categorization\u003c/p\u003e","description":"","filename":"FigS4Permutationposcontr.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/6868169da7c21d3fdd7bdb37.pdf"},{"id":58222971,"identity":"fb4aed09-43d0-4d8a-82f8-3400bc343eb1","added_by":"auto","created_at":"2024-06-12 17:09:26","extension":"pdf","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":15993,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Figure S5: SBS1-score is correlated with chronological age.\u003c/p\u003e","description":"","filename":"FigS5SBS1MSIstat.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/83fb4d9b62bda7ef28b28c00.pdf"},{"id":58222973,"identity":"30099aa0-2720-4323-9078-22cab6d9e8f3","added_by":"auto","created_at":"2024-06-12 17:09:26","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":1819198,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eSupplementary Tables S1–S5\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThis workbook contains 5 Supplementary tables: Table S1 shows an overview of the included CpGs per exposome trait per marker selection threshold; Table S2 provides the CpGs and their respective weights per trait; Table S3 presents the results obtained in the discovery phase (TCGA-COAD) for all thresholds; Table S4 presents the results obtained in the replication meta-analysis CRC; and Table S5 presents the results obtained in TCGA-READ and the meta-analysis of colon and rectal cancer datasets.\u003c/p\u003e","description":"","filename":"SupplementarytablesEpigeneticFingerprintsLinkEarlyOnsetColonandRectalCancertoPesticideExposure04062024.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-4528579/v1/c5e95a39771c49c9c8b86619.xlsx"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nIB has received accommodation and travel expenses from Amgen, Merck, Sanofi, and Servier; and personal speaker honoraria from Astra Zeneca. EE has received personal honoraria from Amgen, Bayer, BMS, Boehringer Ingelheim, Cure Teq AG, Hoffman La – Roche, Janssen, Lilly, Medscape, Merck Serono, MSD, Novartis, Organon, Pfizer, Pierre Fabre, Repare Therapeutics Inc., RIN Institute Inc., Sanofi, Seagen International, GmbH, Servier, and Takeda. JT reports personal financial interest in the form of scientific consultancy role for Alentis Therapeutics, AstraZeneca, Aveo Oncology, Boehringer Ingelheim, Cardiff Oncology, CARSgen Therapeutics, Chugai, Daiichi Sankyo, F. Hoffmann-La Roche Ltd, Genentech Inc, hC Bioscience, Ikena Oncology, Immodulon Therapeutics, Inspirna Inc, Lilly, Menarini, Merck Serono, Merus, MSD, Mirati, Neophore, Novartis, Ona Therapeutics, Orion Biotechnology, Peptomyc, Pfizer, Pierre Fabre, Samsung Bioepis, Sanofi, Scandion Oncology, Scorpion Therapeutics, Seattle Genetics, Servier, Sotio Biotech, Taiho, Takeda Oncology and Tolremo Therapeutics. Stocks: Oniria Therapeutics, Alentis Therapeutics, Pangaea Oncology and 1TRIALSP, and also an educational collaboration with Medscape Education, PeerView Institute for Medical Education and Physicians Education Resource (PER). JAS, OBI, and SCEM declare no competing interests.","formattedTitle":"Epigenetic Fingerprints Link Early-Onset Colon and Rectal Cancer to Pesticide Exposure","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGlobally, colorectal cancer (CRC) ranks as the third leading cancer type and the second most common cause of cancer-related death (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). As CRC is an aging-associated disease, the rates of CRC incidence and death grow steadily with age, with an estimated 90% of worldwide cases and deaths occurring in individuals over 50 years old (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). However, in recent years, worldwide cancer registries have reported a disproportionate uptick in the incidence of early-onset CRC (EOCRC), generally defined as CRC diagnosed in individuals younger than 50 years of age (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). In the U.S., the average annual percent change in the incidence of EOCRC was 2.2% between 2008 and 2012, while similar trends have been observed in New Zealand, Canada, and Europe (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). This tendency is expected to continue in the coming years, with a predicted increase in the incidence of rectal and colon cancer in patients below the age of 34 by 2030 of 124% and 90%, respectively (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eEOCRC presents unique clinical and pathological characteristics in comparison with CRC in older patients, with a predominance of rectal and left colon tumors, higher prevalence of synchronous and metachronous CRC, higher frequency of metastatic disease at diagnosis, and more aggressive and less differentiated tumors (\u003cspan additionalcitationids=\"CR8 CR9\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). However, at the genome level, molecular alterations are similar to those described in older patients, and no specific alterations have been described (\u003cspan additionalcitationids=\"CR12 CR13 CR14\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). Concerning epigenetic changes, hypomethylation of long-interspersed nucleotide elements-1 (LINE-1), which acts as a surrogate marker of global DNA methylation status in the genome, has been reported in patients with EOCRC (\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e), while an assortment of lifestyle and environmental exposures have been associated with lower LINE-1 methylation levels in blood cells in the general population (\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAs our understanding of risk factors for CRC expands, it is becoming increasingly evident that lifestyle factors, such as diet and exercise, can significantly influence CRC risk. Despite this growing knowledge, the reasons behind the rising incidence of CRC in young patients remain unclear. Cancer predisposition syndromes are found to be the underlying cause in only 10\u0026ndash;30% of patients with EOCRC, hence most patients do not present any genetic predisposition (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). Emerging hypotheses suggest that lifestyle and environmental exposures \u0026mdash; collectively known as the exposome \u0026mdash; may play a critical role. However, efforts to identify modifiable risk factors specific to EOCRC have met with only partial success (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). Most studies compare early-onset patients with controls, primarily identifying the same risk factors as those for later-onset cases (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). The main investigated modifiable traits that are associated with EOCRC are smoking habits, diet, educational attainment, Body Mass Index (BMI), and microbiome (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). Concerning microbiome, patients with CRC present a distinct microbiome composition in comparison with healthy subjects. Although specific differences in the microbiome between EOCRC and those diagnosed at later ages remain unexplored, E. coli strains harboring the polyketide synthase (pks) operon, producing the genotoxic compound colibactin and described as contributors to CRC, have been found in CRC cases diagnosed at a younger age (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOne possible reason for this limited success is the lack of quantitative measurements of exposome traits in cancer cohorts, making it challenging to establish clear associations with cancer risk. However, emerging studies have identified exposome-induced changes in CpG site-specific DNA methylation levels (\u003cspan additionalcitationids=\"CR29 CR30 CR31 CR32 CR33 CR34 CR35 CR36 CR37 CR38 CR39 CR40 CR41 CR42\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e), providing a new avenue for investigation. As DNA methylation data is often available in cancer cohorts, utilizing exposome-related methylation changes could serve as a valuable proxy for direct exposome measurements. This approach may enhance our ability to pinpoint specific exposome risk factors and improve our understanding of their role in EOCRC development.\u003c/p\u003e \u003cp\u003eIn this study, we aimed to explore the exposome traits that could be contributing to the development of EOCRC compared to later-onset CRC (LOCRC), here defined as CRC diagnosed in individuals 70 years and older, using epigenetic markers. To this end, we constructed methylation risk scores (MRS) for lifestyle and environmental factors using DNA methylation data from The Cancer Genome Atlas (TCGA) as discovery and six replication datasets and compared the constructed MRSs between EOCRC and LOCRC. Furthermore, we investigated the use of single base substitution (SBS) mutational signature to account for tumor age rather than the patients\u0026rsquo; age at diagnosis. The relationship between pesticide use intensity and EOCRC incidence in the U.S. was further investigated using population-based data.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cem\u003eCohort Description\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eFor the discovery phase of this study, we utilized colon adenocarcinoma (COAD) samples obtained from TCGA (44). The subsequent replication phase was carried out through a meta-analysis, incorporating data from studies on colon cancer (GSE131013 (45) and GSE42752 (46)), rectal cancer (TCGA-READ and GSE39958 (47)), and colorectal cancer (GSE101764 (48) and GSE77954 (49)). We categorized the participants by age: those diagnosed before the age of 50 were identified as early-onset cases, while those diagnosed at 70 or older were designated as later-onset CRC (LOCRC) cases. The population characteristics across all datasets are presented in Table 1.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eTable 1: Study population Characteristics\u003c/em\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"607\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eStudy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eN total\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eFemales, N (%)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eAge, years mean (SD)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eCancer type\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eDiscovery dataset:\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eTCGA-COAD\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"top\"\u003e\u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"top\"\u003e\u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"top\"\u003e\u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eColon\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eEarly-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e18 (58)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e43 (5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eLater-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e100\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e43 (43)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e78 (5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eReplication datasets:\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cstrong\u003eTCGA-READ\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eRectal\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eEarly-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e6 (43)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e44 (6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eLater-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e21 (70)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e76 (6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cstrong\u003eGSE39958\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eRectal\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eEarly-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e3 (25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e45 (7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eLater-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e2 (22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e74 (5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cstrong\u003eGSE42752\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eColon\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eEarly-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e3 (75)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e46 (3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eLater-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e5 (71)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e76 (4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cstrong\u003eGSE77954\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eColorectal\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eEarly-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e3 (100)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e48 (1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eLater-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e3 (30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e77 (6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cstrong\u003eGSE101764\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eColorectal\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eEarly-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e3 (23)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e38 (8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eLater-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e13 (35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e76 (5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cstrong\u003eGSE131013\u003csup\u003e*\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eColon\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eEarly-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e0 (0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e46 (4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"25.657894736842106%\"\u003e\n \u003cp\u003e\u003cem\u003eLater-onset\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"12.664473684210526%\" valign=\"bottom\"\u003e\n \u003cp\u003e58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"18.914473684210527%\" valign=\"bottom\"\u003e\n \u003cp\u003e13 (22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"25.32894736842105%\" valign=\"bottom\"\u003e\n \u003cp\u003e76 (5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"17.43421052631579%\" colspan=\"2\" valign=\"top\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"99.67051070840198%\" colspan=\"5\" valign=\"top\"\u003eEarly-onset cases are defined as participants younger than 50 years, while later-onset cases comprise\u003cstrong\u003e\u003cbr\u003e \u003c/strong\u003eparticipants diagnosed at age 70 years and older. Dataset GSE77954 encompasses primary (n=7) and \u003cstrong\u003e\u003cbr\u003e \u003c/strong\u003emetastatic samples (n=6). \u003csup\u003e*\u003c/sup\u003eNotably, the early-onset category in\u003csup\u003e \u003c/sup\u003eGSE131013 comprises solely male participants while GSE77954 includes only female participants. Consequently, adjustment for sex was not feasible in these two datasets.\u003cbr\u003e\u003c/td\u003e\n \u003ctd width=\"0.32948929159802304%\"\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eExposome-related DNA methylation marker sets\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eIn exploring the exposome\u0026apos;s influence on early-onset versus later-onset colon and rectal cancers, our research focused on a curated list of 29 (28\u0026ndash;43) lifestyle and environmental factors. The analyzed traits encompassed 10 lifestyle factors: the Alternative Healthy Eating Index (AHEI) (28), alcohol consumption (29), birth weight (30), BMI (continuous variable in kg/m\u003csup\u003e2\u003c/sup\u003e) (31), coffee consumption (32), education level (33), Mediterranean Diet Score (MDS) (28), obesity (defined as \u0026ge;30 kg/m\u003csup\u003e2\u003c/sup\u003e) (34), smoking habits (35), and smoking inference model (smoking-Maas) (36). Furthermore, we examined 5 air pollution particles: nitrogen dioxide (NO\u003csub\u003e2\u003c/sub\u003e)\u003csub\u003e\u0026nbsp;\u003c/sub\u003e(37\u0026ndash;39), polychlorinated biphenyls (PCBs) (42), and particulate matter (PM) \u0026lt;10 micrometers (\u0026micro;m) in diameter (PM\u003csub\u003e10\u003c/sub\u003e)\u003csub\u003e\u0026nbsp;\u003c/sub\u003e(38\u0026ndash;40), \u0026lt;2.5 \u0026micro;m (PM\u003csub\u003e2.5\u003c/sub\u003e) (37\u0026ndash;41), and between 2.5 and 10 \u0026micro;m (PM\u003csub\u003e2.5-10\u003c/sub\u003e)\u003csub\u003e\u0026nbsp;\u003c/sub\u003e(40). In addition, we included 14 pesticides encompassing 2,4-dichlorophenoxyacetic acid (2,4-D), atrazine, acetochlor, chlordane, dicamba, malathion, Dichlorodiphenyltrichloroethane (DDT), heptachlor, lindane, glyphosate, mesotrione, metolachlor, picloram, and toxaphene (43). For the marker selection, we identified for each trait significantly associated CpG sites from extensive epigenome-wide association studies (EWAS), employing various significance thresholds, namely P\u0026lt;1.2\u0026times;10\u003csup\u003e-7\u003c/sup\u003e, P\u0026lt;1.0\u0026times;10\u003csup\u003e-5\u003c/sup\u003e, and false discovery rates (FDR) of \u0026lt;0.01, \u0026lt;0.05, and \u0026lt;0.1.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eExposome-related methylation risk scores\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eUtilizing available EWAS summary statistics of each trait for the five marker selection thresholds, we identified 73 exposome CpG sets across the 29 exposome traits. These sets were used to compute 73 weighted methylation risk scores (MRSs), utilizing DNA methylation beta-values adjusted for epigenetic age estimators derived from the Horvath clock (50). The number of CpG sites associated with each trait and their respective significance thresholds can be found in Supplementary Table S1 and their corresponding weights in Supplementary Table S2. To elucidate the exposome impact on early-onset colon and rectal cancer cases, we compared the 73 MRSs across early- vs. later-onset (reference group) patients, using multivariate logistic regression models. As there are sex disparities in CRC incidence (51), we adjusted the regressions for sex, when possible. In the discovery dataset, positive associations were observed for MRSs related to PCB, PM\u003csub\u003e10\u003c/sub\u003e, the smoking-Maas model, heptachlor, metolachlor, picloram, and toxaphene, while negative associations were found for MRSs corresponding to BMI, education level, MDS, obesity, atrazine, malathion, and mesotrione (Fig. 1 and Supplementary Table S3).\u003c/p\u003e\n\u003cp\u003eWe highlight the data for four lifestyle factors previously linked to colon and rectal cancers, including the MDS (52) (Fig. 2a) and education level (53) (Fig. 2b), which are considered protective factors, as well as smoking habits (54) (Fig. 2c) and obesity status (55) (Fig. 2d), which are recognized as risk factors. To elucidate the directionality of our findings, the heatmaps in the left panels of Fig. 2 show the methylation level distributions across CpGs featured in each of the four MRSs, their direction in the original EWAS and sorted by the derived MRSs. The heatmaps depict that an increased MRS correlates with higher beta-values in CpGs with positive associations in the EWAS, and lower beta-values in CpGs with negative associations (please refer to Supplementary Fig. S1 for a more detailed explanation). These results suggest that an elevated MRS mirrors greater exposure levels in the original EWAS. Specifically, for patients with early-onset colon cancer, this suggests deviations from the MDS (Padj. = 0.037) (Fig. 2a), lower education levels (Padj. = 0.025) (Fig. 2b), increased smoking exposure (Padj. = 0.010) (Fig. 2c), and lower obesity rates (Padj. = 0.011) (Fig. 2d) in comparison to those with later-onset, as illustrated in the middle panels of Fig. 2. The association of lower obesity rates in early-onset cases was verified utilizing physical metrics from TCGA-COAD. Colon cancer patients with a BMI over 30kg/m\u003csup\u003e2\u003c/sup\u003e, as measured in the clinic, were categorized as obese, resulting in 4 out of 24 early-onset and 18 out of 72 later-onset patients marked as obese. This provides a relative risk (RR) of 0.67 (95% CI: 0.26-1.76) for obesity in early-onset colon cancer patients within the TCGA-COAD cohort (Supplementary Fig. S2), supporting the results obtained with the MRS for obesity, and validating our methodology.\u003c/p\u003e\n\u003cp\u003eExtending our investigation, we conducted a meta-analysis of the six replication datasets (Fig. 1 and Fig. 2 right panels, and Supplementary Table S4). This analysis corroborated the initial findings, notably the associations of non-adherence to MDS (P = 0.011, Padj. = 0.051), lower educational levels (P = 0.0039, Padj. = 0.025), and higher smoking exposure (P = 0.0025, Padj. = 0.024) in EOCRC (Fig. 2, right panel). Moreover, we conducted separate meta-analyses of the datasets comprising only rectal cancer samples (TCGA-READ and GSE39958) and only colon cancer samples (GSE131013 and GSE42752). The results obtained in TCGA-READ and the meta-analyses are presented in Supplementary Fig. S3 and Supplementary Table S5.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePicloram-related methylation risk scores\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eOur results highlight a novel association between the MRSs for pesticide picloram and the incidence of early-onset colon and rectal cancer, in comparison to later-onset cases, in both the discovery and meta-analysis (Fig. 1). We explored the directionality of these results further, highlighting the data for the MRS employing the genome-wide marker selection threshold (MRS-GW). We observed that a higher exposure level, as indicated by the original EWAS direction, is associated with an elevated MRS (Fig. 3a). This association highlights an augmented exposure to picloram among patients with early-onset colon cancer (Padj. = 0.00049) (Fig. 3b), a finding consistently supported by our meta-analysis (P = 0.021; Padj. = 0.081; OR: 1.6 [95% CI: 1.07-2.38]) (Fig. 3c).\u003c/p\u003e\n\u003cp\u003eTo ascertain the reliability of our findings, we executed two distinct permutation tests to ensure that the observed associations stem from biological relationships rather than being artifacts of particular CpG selections or patient classifications. Initially, the examination of the CpG sites in our MRSs revealed that the CpG sites from the picloram MRS-GW were 13\u003csup\u003eth\u003c/sup\u003e in terms of significance among 10,000 permutations (Fig. 3d). Furthermore, patient classification permutation identified age-based classification as the second most significant, based on picloram MRS-GW, among 1,000 permutations, as depicted in Fig. 3d. The outcomes of the CpG site permutations for MDS, education levels, the smoking-Maas model, and obesity are detailed in Supplementary Fig. S4a, while the findings from the onset categorization permutations are shown in Supplementary Fig. S4b.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eYoung tumors associated with picloram exposure\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eCurrent patient classification of CRC into early-onset or later-onset categories relies on the patient\u0026apos;s age at diagnosis. However, this method is flawed, as the interval between tumor initiation and diagnosis varies significantly among patients, rendering age at diagnosis an unreliable indicator of the tumor\u0026apos;s actual age. To address this, we assessed if the single-base substitution signature 1 (SBS1) score, an indicator of the number of mitotic divisions a cell has undergone (56,57), can instead be used as tumor age. For this purpose, we selected patients from TCGA-COAD with data available on DNA methylation and mutational signatures. Furthermore, we excluded patients exhibiting microsatellite instability (MSI), considering MSI arises from defective DNA mismatch repair, inducing distinct mutational patterns that might drive tumorigenesis through mechanisms different from those in microsatellite stable (MSS) tumors (Supplementary Fig. S5) (58,59). The distribution of SBS1 mutations across different age groups\u0026mdash; early-onset, middle-onset (aged between 50 and 69 years), and later-onset\u0026mdash; among the 173 patients included in the study is detailed in Fig. 3e. Upon comparing the early-onset cases (N = 25) against later-onset cases (N = 72) and the picloram MRS-GW, we observed a statistically significant difference (OR: 2.99 [95% CI: 1.70-5.85]; P = 4.27\u0026times;10\u003csup\u003e-4\u003c/sup\u003e). Next, we employed a SBS1 score threshold, identifying 72 young (SBS1\u0026lt;60) and 101 old tumors (SBS1\u0026ge;60). The novel patient categorization underscored a significant association with picloram MRS-GW (OR: 1.84 [95% CI: 1.31- 2.66]; P = 6.57\u0026times;10\u003csup\u003e-4\u003c/sup\u003e) (Fig. 3f). The chronological age distribution for these SBS1-categorized young and old tumors is provided in Fig. 3g.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e\u0026nbsp;Pesticide use and EOCRC incidence in population data\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe use of MRSs as a proxy for pesticide exposure identified significant associations with several pesticides. Next, we aimed to validate the obtained results employing pesticide use intensity and incidence of EOCRC for available overlapping counties in California, Connecticut, Georgia, Iowa, New Mexico, Utah, and Washington in the United States. Based on data availability, we extracted pesticide usage from the Pesticide National Synthesis Project encompassing acetochlor, 2,4-D, atrazine, dicamba, glyphosate, mesotrione, and picloram. The EOCRC incidence rates were extracted from the Surveillance, Epidemiology, and End Results (SEER), encompassing EOCRC rates measured in 8 registries from 1975 to 2020 (SEER8) or in 12 registries measured from 1992 to 2020 (SEER12). The total number of included observations, the number of measured years times the number of overlapping counties, for acetochlor (SEER8: N\u003csub\u003e\u0026nbsp;\u003c/sub\u003e= 1.111, SEER12: N = 1.196), 2,4-D (N = 1.983, N = 2.059), atrazine (N = 1.871, N = 1890), dicamba (N = 1.909, N = 1.964), glyphosate (N = 2.002, N = 2.097), mesotrione (N = 636, N = 636), and picloram (N = 1.531, N = 1.548) are depicted in Fig. 4a. Specifically, Fig. 4b shows the average (log) picloram use intensity and the average EOCRC incidence rates between 1992 and 2012 in the state of Iowa. To assess the relationship between pesticide-use intensity (exposure) and age-adjusted EOCRC incidence rates, we utilized linear mixed models adjusting for the years of data collection and a random effect to accommodate county-level variations. Our approach also tested for interaction effects between pesticide use intensity and the years of data collection, which proved to be non-significant for all pesticides under study (data not shown). Significant associations were found between the pesticide use intensity of multiple pesticides and EOCRC incidence in both SEER8 and SEER12, including glyphosate (SEER 8; P = 1.18\u0026times;10\u003csup\u003e-5\u003c/sup\u003e, SEER 12; P = 2.02\u0026times;10\u003csup\u003e-4\u003c/sup\u003e), atrazine (P = 1.81\u0026times;10\u003csup\u003e-4\u003c/sup\u003e, P = 4.21\u0026times;10\u003csup\u003e-3\u003c/sup\u003e), picloram (P = 2.87\u0026times;10\u003csup\u003e-3\u003c/sup\u003e, P = 1.82\u0026times;10\u003csup\u003e-2\u003c/sup\u003e), 2,4-D (P = 4.16\u0026times;10\u003csup\u003e-3\u003c/sup\u003e, P = 1.8\u0026times;10\u003csup\u003e-3\u003c/sup\u003e), and dicamba (P = 4.94\u0026times;10\u003csup\u003e-3\u003c/sup\u003e, P = 4.10\u0026times;10\u003csup\u003e-2\u003c/sup\u003e), depicted in Fig. 4c.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we explored the exposome traits that could be specifically contributing to EOCRC disease through the use of epigenetic markers. There is growing evidence of a worldwide rise in the incidence of CRC cases in individuals younger than 50, while the number of cases in later ages is decreasing in most countries (\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e). Epidemiological studies have linked CRC to non-modifiable factors, including sex, age, ethnicity, and genetic predisposition (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e, \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e), and to external exposures such as adiposity, smoking and drinking status, household income, Western pattern diet, and physical inactivity (\u003cspan additionalcitationids=\"CR64 CR65 CR66 CR67 CR68 CR69 CR70\" citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e). However, these exposures have not been specifically associated with EOCRC, leaving the reasons for the rising EOCRC incidence rates unclear. To the best of our knowledge, this is the first research addressing the exposome impact \u0026mdash; including environmental factors and lifestyle \u0026mdash; on early-onset colon and rectal cancer cases through epigenetic fingerprints.\u003c/p\u003e \u003cp\u003eThe effects of environmental exposures such as microplastics and nanoplastics (MNPs), air pollutants, and other potential carcinogenic agents on human health and CRC are just beginning to be documented (\u003cspan additionalcitationids=\"CR71 CR72 CR73 CR74 CR75 CR76\" citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e). Exposures through lifespan may change, so the analyses in terms of both categorical and quantitative measurement are challenging and may raise concerns about potential spurious associations. Our results consistently link EOCRC to exposure to the pesticide picloram based on epigenetic fingerprints. Picloram\u0026rsquo;s mechanism of action as an herbicide relies on its capacity to mimic the plant growth hormones auxins and to inhibit the enzymes that break down auxins, which leads to more persistent effects than the natural hormone (\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e). Therefore, picloram disrupts normal growth, causing abnormal stimulation and maturation of tissues, which triggers growth discontinuation, root deterioration and eventually plant death. Picloram was first registered as a pesticide in the U.S. in 1964 and the herbicide and its derivatives have generally shown to be of moderate to low acute toxicity in laboratory animals. However, dietary exposure to residues of picloram is plausible as it has been found in grain and meat by-products, and the effects of long-term use on human health have not been described so far. If the use of picloram in crops started in the mid and late 20th century, then-current individuals with LOCRC were not exposed to it during their childhood, while cases of EOCRC were and have been for a longer part of their lives, which could explain our results. The association between pesticide exposure and EOCRC using population-based data further validates the association with picloram. Besides picloram, we also show evidence for associations between EOCRC and exposure to glyphosate and atrazine. Glyphosate is already categorized as \u0026ldquo;probably carcinogenic to humans\u0026rdquo; by the International Agency for Research on Cancer (IARC), suggesting the validity of the obtained results (\u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e). Although atrazine has been banned in the European Union since 2004, the U.S. EPA still approves its continued use (\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e). Already during the IARC review in 1999, sufficient evidence confirmed the carcinogenicity of atrazine in experimental animals, but as evidence showed that the identified mechanism in rats was not relevant to humans, atrazine was categorized in group 3 (\u0026ldquo;not classifiable as to its carcinogenicity in humans\u0026rdquo;). Novel studies have identified alternative mechanisms for how atrazine could cause carcinogenesis in humans, e.g. by damaging DNA integrity, the stability of the cell genome, DNA double-strand breaks, and the activation of DNA damage checkpoints (\u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e). This resulted in the Monographs program for 2025\u0026ndash;29 (\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e) to warrant a high-priority re-evaluation of the IACR classification as a result of the identified new human cancer and mechanistic evidence. Furthermore, a recent update study in the Agricultural Health Study Cohort further suggested the link between atrazine use and several cancer types, including in patients\u0026thinsp;\u0026lt;\u0026thinsp;50 years of age (\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eLower education levels, increased smoking exposure, non-adherence to MDS, and lower obesity rates for patients with EOCRC were found. This suggests that obesity could be a greater contributing factor in LOCRC than in EOCRC. Obesity has been largely described as a risk factor for CRC, although contradictions about its role in tumor aggressiveness, CRC progression, and survival have been reported. For EOCRC, a meta-analysis showed that a BMI greater than 30 kg/m\u003csup\u003e2\u003c/sup\u003e was significantly associated with the development of the disease (\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e) as compared to controls, but considerable heterogeneity among risk estimates was found. Conversely, when comparing EOCRC with patients diagnosed at a later age, older patients were more likely to be obese or overweight (\u003cspan additionalcitationids=\"CR86\" citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e). Of note, in our study, we lack data about longitudinal BMI throughout the life course, but longer exposure through years to high BMI and adiposity in patients with LOCRC could be mediating this effect. It should also be considered that weight loss is a common symptom of CRC, especially in patients with advanced disease, which is, in general, more frequent in EOCRC. Further studies including longitudinal follow up are needed to confirm the role of obesity in EOCRC.\u003c/p\u003e \u003cp\u003eThe patient's age at diagnosis is currently used as the tumor's age, possibly introducing bias in identifying early-onset patients. To address this, we assessed if tumor age can be established by identifying molecular characteristics of aged tumors, such as mutational signatures (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e). Characteristic mutational signatures in cancer genomes arise from various mutational processes, including defects in DNA maintenance mechanisms and both external and internal exposures (\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e). Mutations attributed to the single-base substitution signature 1 (SBS1) are thought to occur during DNA replication in mitosis, suggesting that the rate of SBS1 mutations may serve as an indicator of the number of mitotic divisions a cell has undergone (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e). Furthermore, a correlation exists between the number of SBS1-attributable mutations within a tumor and the patient's age at cancer diagnosis (\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e). Here, we demonstrate that utilizing an SBS1 score threshold to distinguish between younger and older tumors is associated with picloram MRS. This finding suggests that the differences observed may be attributable to the biological age of the tumor rather than the chronological age of the patient.\u003c/p\u003e \u003cp\u003eWe acknowledge the limitations of the current study. First, the number of EOCRC cases in the datasets is relatively small. Second, we recognize low ethnical diversity in the investigation, as TCGA is mostly enriched in non-Hispanic white ethnicity. Although the other datasets consist of patients from Russia, Spain, Germany, South Korea, and the United States, independent studies on other populations are highly encouraged to capture the diversity of worldwide patients with EOCRC in terms of ethnicity, cultural, and socioeconomic differences.\u003c/p\u003e \u003cp\u003eThe study also presents unprecedented strengths. First, the use of MRS as a proxy for exposome factors is a novel approach that allows the exploration of traits that would otherwise not be possible due to limited data availability. Similarly, the lack of longitudinal follow-up measurements or the use of self-reported data can be overcome as MRSs integrate the cumulative impact of the exposure over time. Nevertheless, prospective birth cohorts with long-term follow-up, quantitative exposure measurement, and biomarker and omics analyses throughout life can elucidate the etiology of EOCRC. Plus, given that exposures in early life may be key for the development of CRC, prospective birth cohorts are needed. However, inherent limitations of these studies, such as cost, cohort size, healthy volunteer bias, and long follow-up needed for cases to appear and draw conclusions, should also be considered. Second, several independent cohorts were used, permutation tests were implemented, and the picloram MRS was consistently associated with EOCRC compared to LOCRC. Also, the validation of the association between picloram use intensity and EOCRC in population-level data further strengthens the evidence of picloram as a novel identified risk factor. Third, the exposome, namely the totality of exposures including, among others, diet, lifestyle, and environment, during early life and young adulthood, has changed considerably in the last decades. The categorization of the population into two cohorts with extreme ages (\u0026lt;\u0026thinsp;50 vs\u0026thinsp;\u0026ge;\u0026thinsp;70) and the exclusion of those cases with intermediate age favor the identification of substantial generational changes in the exposome of the two cohorts, while those in between may present a gradual change in their exposure. The relevance and impact of this investigation may be multiple, as it suggests research priorities for primary preventive interventions aimed at behavior modifications and secondary prevention in those individuals exposed to risk factors for EOCRC for current and future generations. Also, it might guide the development of health policies for environmental exposures and regulatory policies for agricultural products. Last, the spotlight on the tumor age rather than on the patient is a novel perspective for CRC research and epidemiology for both EOCRC and those diagnosed over the age of 50, and this could be one of the potential explanations for why no differences in tumor biology between EOCRC and LOCRC have been found so far.\u003c/p\u003e \u003cp\u003eIn conclusion, our findings not only provide exposome traits based on epigenetic fingerprints that could be contributing to the development of CRC, specifically in EOCRC, but also pave the way with a compelling rationale for addressing lifestyle and environmental exposures to mitigate EOCRC risk, highlighting the importance of both personal and policy-level interventions.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cem\u003eStudy population\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe discovery phase of our study utilized colon adenocarcinoma (COAD) samples from TCGA. To ensure the integrity of our analysis, patients with Lynch syndrome (TCGA-A6-6781, TCGA-CM-6674, and TCGA-D5-6927) were excluded. Our replication effort encompassed six datasets, focusing on rectal cancer (TCGA-READ and GSE39958 (47)), colon cancer (GSE131013 (45) and GSE42752 (46)), and colorectal cancer studies (GSE101764 (48) and GSE77954 (49)). For the TCGA samples, we implemented exclusion criteria, removing cases annotated with \"Item in special subset\", \"History of unacceptable prior treatment related to a prior/other malignancy\", \"Case submitted is found to be a recurrence after submission\", \"Neoadjuvant therapy\", \"Synchronous malignancy\", and \"Pathology outside specification\". This led to the exclusion of 10 patients from the COAD dataset and 8 from the READ dataset. Additionally, samples preserved in formalin-fixed paraffin-embedded (FFPE) form were excluded to maintain consistency in sample quality. Duplicate samples were identified and removed based on their plate number, further refining our cohorts for analysis. \u003c/p\u003e\n\u003cp\u003e\u003cem\u003eDNA methylation data processing\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eDNA methylation data was in all datasets obtained using the Illumina Infinium HumanMethylation450 BeadChip. TCGA methylation data was acquired from the Genomic Data Commons (GDC) Data Portal through the \u003cem\u003eTCGABiolinks\u003c/em\u003e package v2.25.0 (88–90). For the additional datasets, idat files or signal intensity files were downloaded from the Gene Expression Omnibus (GEO) database via the \u003cem\u003eGEOquery\u003c/em\u003e package v2.60.0 (91). We processed the tumor sample methylation data with the \u003cem\u003esesame\u003c/em\u003e package v1.19.7 (92–95) in R version 4.3.1. CpG sites showing more than 50% missing data were removed from each dataset. The remaining missing values underwent imputation using the \u003cem\u003eimpute \u003c/em\u003epackage v1.66.0 (96), followed by mean imputation for CpGs with imputed values of 0. DNA methylation beta-values were then converted to M-values using the \u003cem\u003elumi\u003c/em\u003e package v2.44.0 (97–100). Next, we estimated the epigenetic age using the Horvath clock, following the R software tutorial (50), and adjusted the DNA methylation m-values for the estimated age, to account for the influence of age on methylation. Finally, we converted the m-values back to beta-values for further analysis. The final dataset refinement involved excluding CpGs that were not present across all datasets, those associated with cross-reactive probes, and CpGs identified as single nucleotide polymorphisms (101).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eExposome-related traits marker selection\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe conducted a detailed literature review to compile a list of exposome traits potentially influencing the risk of early-onset colon and rectal cancer, ensuring the inclusion of traits with available comprehensive epigenome-wide association studies (EWAS). This process led to the identification of 29 key exposome traits (28–43). To address the challenge of smaller sample sizes in air pollution-related EWAS, we combined CpG sites from multiple studies. The selected exposome traits span a wide range, including lifestyle factors, environmental exposures, and a selection of pesticides. CpG sites were chosen using stringent significance criteria: P\u0026lt;1.2×10\u003csup\u003e-7\u003c/sup\u003e, P\u0026lt;1.0×10\u003csup\u003e-5\u003c/sup\u003e, and false discovery rates (FDR) of \u0026lt;0.01, \u0026lt;0.05, and \u0026lt;0.1. We employed a tiered strategy for marker selection to balance sensitivity and specificity, facilitating a comprehensive assessment of the relationship between exposome exposures, DNA methylation levels, and early-onset colon and rectal cancer. This approach includes a broad range of CpG sites explored using less stringent p-value thresholds (e.g., P\u0026lt;1.0×10\u003csup\u003e-5\u003c/sup\u003e and FDR\u0026lt;0.1), increasing sensitivity to include CpG sites with potential biological relationships. This was instrumental in selecting both highly confident associations and those of potential biological relevance that may not meet the strictest statistical thresholds. In addition, we included CpG sites using more stringent criteria (e.g., P\u0026lt;1.2×10\u003csup\u003e-7\u003c/sup\u003e, FDR\u0026lt;0.05, and FDR\u0026lt;0.01), enhancing specificity and reducing the risk of including false positive CpG sites. This approach allowed for a nuanced exploration of the data, ensuring that both robust and biologically meaningful associated CpG sites were included in our analysis.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eExposome-related methylation risk scores in early- vs later-onset patients \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo assess the impact of these exposome traits on early-onset colon and rectal cancer risk, we computed individual methylation risk scores (MRS) by applying effect sizes from relevant EWAS to methylation levels at each CpG site and summing these across all pertinent CpGs. The obtained scores were normalized to a mean of 0 and a standard deviation of 1, exclusively utilizing the data from EOCRC and LOCRC patients.\u003c/p\u003e\n\u003cp\u003eWe examined the correlation between these MRSs and the incidence of early-onset cancer, comparing them to later-onset patients using logistic regression analyses adjusted for sex using the \u003cem\u003estats \u003c/em\u003ev4.1.3. R package. This analysis spanned the discovery dataset TCGA-COAD and extended to six validation cohorts. Models were excluded from the subsequent meta-analysis if they issued fitting warnings or were unable to estimate the 95% confidence interval. The collective data from the validation cohorts were subjected to a meta-analysis utilizing the \u003cem\u003emetafor\u003c/em\u003e R package v4.4-0 (102). The obtained p-values were FDR (103) corrected for the number of traits in each marker selection threshold, including 19 traits in GW, 26 in P1E5, 9 in F01, 11 in F005, and 8 in F001, using the p.adjust function in \u003cem\u003estats\u003c/em\u003e v4.1.3. The results were then visually summarized in forest plots, offering a clear depiction of the associations between MRS and the risk of EOCRC relative to later-onset cases. \u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePermutation analysis for CpG selection and patient categorization\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the robustness of our findings, we conducted two permutation analyses, ensuring that the observed associations are not artifacts of the specific CpG selection or patient grouping but reflect underlying biological relationships. The first analysis involved randomly assigning patients to the MRSs. Specifically, we shuffled the early- and later-onset colon cancer patients in the TCGA-COAD dataset 1,000 times, creating alternative configurations of the dataset. For each rearranged dataset, we performed logistic regression analyses, adjusting for sex, to assess the stability of the association between MRSs and early-onset across these permutations. The second analysis targeted the specific assignment of CpGs to their corresponding effect sizes (weights). We temporarily removed the CpGs that were initially selected for each exposome trait to create a pool of potential CpGs. From this pool, we generated 10,000 new sets, each containing an equal number of CpGs as in the original analysis but selected randomly. These randomly chosen CpGs were then used to construct new MRSs for each trait. Subsequently, logistic regression analyses, adjusted for sex, were applied to these permutation-derived MRSs to test the significance of the associations under these randomized conditions.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eYoung vs. older tumors using single-base substitution signature 1 \u003c/em\u003e\u003c/p\u003e\n\u003cp\u003ePatients from the TCGA-COAD dataset were selected when presented with comprehensive profiles including DNA methylation data, mutational signatures (https://dcc.icgc.org/releases/PCAWG), and confirmed microsatellite stability. To ensure data integrity, we excluded extreme outlier patients, identified by SBS1 scores exceeding more than three times the standard deviation above the mean SBS1 score. The patients were categorized as early-onset (aged \u0026lt;50), middle-onset (aged 50 to 69 years), and later-onset (aged ≥70 years) cases. Initially, we assessed the differential impact of picloram MRS-GW between the early- and later-onset groups within this subset. Subsequently, we aimed to enhance the specificity of our findings by integrating the concept of tumor age, inferred from the SBS1 score, into our analysis of early- and later-onset categories. In a novel approach, we explored the efficacy of using tumor age—defined by an SBS1 score of 60 as a threshold—as an alternative to chronological age in distinguishing between younger and older tumors. We selected the threshold of an SBS1 score of 60 as this was the first quartile in later-onset patients and the median in middle-onset patients. The third quartile in early-onset patients was an SBS1 score of 67.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePesticide use and early-onset CRC in the United States\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe results obtained for pesticide MRSs were validated using population data from the U.S. We assessed the association between pesticide use and EOCRC incidence across overlapping counties. Utilizing data from the Pesticide National Synthesis Project, we analyzed county-level pesticide usage between 1992 and 2012 for specific chemicals: acetochlor, 2,4-D, atrazine, dicamba, glyphosate, mesotrione, and picloram. These pesticides were selected based on the availability of complete data. To estimate pesticide use intensity for each county, we divided the total estimated pesticide use by the county's area in square miles.\u003c/p\u003e\n\u003cp\u003eEOCRC incidence rates were obtained from the Surveillance, Epidemiology, and End Results (SEER) database using the SEER*Stat software (version 8.4.1) (104). This comprehensive dataset allowed us to incorporate Research Plus Data spanning from 1975 to 2020 for SEER8 and from 1992 to 2020 for SEER12. We calculated age-adjusted EOCRC incidence rates (for individuals aged 25-49) at the county level annually, normalizing the number of cases per 100,000 population. Counties with no cases for more than 50% of the years were excluded to mitigate the impact of sparse data on our analysis.\u003c/p\u003e\n\u003cp\u003eThe pesticide data and EOCRC incidence data were overlapped at county level. To refine our analysis and reduce the influence of extreme values, we removed the top and bottom 5% measurements based on log-transformed pesticide-use intensity. Our analytical model, a linear mixed model, was employed to examine the relationship between pesticide use intensity and EOCRC incidence, adjusting for the years measured and including a random effect to account for variations across counties. We further explored the temporal dynamics of this association by incorporating an interaction term between pesticide use intensity and the measured years. The significance of the interaction term, indicative of changes in the association over time, was assessed using ANOVA for each pesticide under study. Linear mixed models and ANOVA analyses were executed with the \u003cem\u003elmerTest R \u003c/em\u003epackage v3.1-3 (105). Statistical analyses were performed in R v4.1.3 (106). \u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll methylation and related clinical data used in this study are publicly available via the GDC Data portal (https://portal.gdc.cancer.gov/) for TCGA datasets or via the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/) for GSE131013, GSE42752, GSE39958, GSE101764, and GSE77954. Pesticide use data can be extracted from the National Water-Quality Assessment (NAWQA) project (https://water.usgs.gov/nawqa/pnsp/usage/maps/county-level/). Access to the Research Plus Data from The Surveillance, Epidemiology, and End Results (SEER) cohort can be requested via https://seer.cancer.gov/data/access.html.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode\u003c/strong\u003e \u003cstrong\u003eavailability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe code used for the analysis in this study is publicly available from the following repository: https://github.com/CancerCompBioLab/EOCRCexposome. All the input files needed to replicate our findings and the results obtained during the study are also available from the GitHub repository.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe first and foremost thank the participants and their families included in the used studies. We are grateful to the investigators and data management teams who recruited the participants and to the pathologists who collected the samples. The results shown here are in part based upon data generated by the TCGA Research Network: https://www.cancer.gov/tcga and the Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*Stat Database: Incidence - SEER \u0026nbsp;Plus Research Data, 8 Registries, Nov 2021 Sub (1975-2019) and 12 Registries, Nov 2021 Sub (1992-2019) - Linked To County Attributes based on the November 2021 submission.\u003c/p\u003e\n\u003cp\u003eWe thank Javier Carmona, PhD; and the members of the Cancer Computational Biology Group for the helpful discussions. This work was supported by CMS2022-135428, RYC2019-026576-I, PID2020-115097RA-I00, and ISCIII grant FORT23/00034, and Fundacion \u0026ldquo;la Caixa\u0026rdquo; (to J.A. Seoane), Juan de la Cierva JDC2022-048829-I (to S.C.E. Maas).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAuthors and Affiliations\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eVall d\u0026rsquo;Hebron Institute of Oncology (VHIO), Barcelona, Catalonia, Spain\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eSilvana C.E. Maas, Iosune Baraibar, Odei Blanco Irazuegui, Josep Tabernero, Elena Elez \u0026amp; Jose A. Seoane\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eResearch programme on Biomedical Informatics (GRIB), Hospital del Mar Research Institute, PRBB, Barcelona\u0026nbsp;\u003c/u\u003e\u003c/p\u003e\n\u003cp\u003eOdei Blanco Irazuegui\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eContributions\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eJAS, EE, and SCEM conceived the study design. SCEM and OBI analyzed the data. JAS, EE, JT, SCEM, and IB interpreted the data and results. SCEM, JAS, and IB drafted the paper and SCEM drafted the figures. All authors critically reviewed the paper and the results and approved the final version.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCompeting interests\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eIB has received accommodation and travel expenses from Amgen, Merck, Sanofi, and Servier; and personal speaker honoraria from Astra Zeneca. EE has received personal honoraria from Amgen, Bayer, BMS, Boehringer Ingelheim, Cure Teq AG, Hoffman La \u0026ndash; Roche, Janssen, Lilly, Medscape, Merck Serono, MSD, Novartis, Organon, Pfizer, Pierre Fabre, Repare Therapeutics Inc., RIN Institute Inc., Sanofi, Seagen International, GmbH, Servier, and Takeda. JT reports personal financial interest in the form of scientific consultancy role for Alentis Therapeutics, AstraZeneca, Aveo Oncology, Boehringer Ingelheim, Cardiff Oncology, CARSgen Therapeutics, Chugai,\u0026nbsp;Daiichi Sankyo,\u0026nbsp;F. Hoffmann-La Roche Ltd, Genentech Inc, hC Bioscience, Ikena Oncology, Immodulon Therapeutics, Inspirna Inc, Lilly, Menarini, Merck Serono, Merus, MSD, Mirati, Neophore, Novartis, Ona Therapeutics, Orion Biotechnology, Peptomyc, Pfizer, Pierre Fabre, Samsung Bioepis, Sanofi, Scandion Oncology, Scorpion Therapeutics, Seattle Genetics, Servier, Sotio Biotech, Taiho, Takeda Oncology and Tolremo Therapeutics. Stocks: Oniria Therapeutics, Alentis Therapeutics, Pangaea Oncology and 1TRIALSP, and also an educational collaboration with Medscape Education, PeerView Institute for Medical Education and Physicians Education Resource (PER). JAS, OBI, and SCEM declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024 Apr 4;74(3):229\u0026ndash;63.\u003c/li\u003e\n\u003cli\u003eSiegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin. 2024 Jan 17;74(1):12\u0026ndash;49.\u003c/li\u003e\n\u003cli\u003eSiegel RL, Torre LA, Soerjomataram I, Hayes RB, Bray F, Weber TK, et al. Global patterns and trends in colorectal cancer incidence in young adults. Gut. 2019 Dec;68(12):2179\u0026ndash;85.\u003c/li\u003e\n\u003cli\u003eLui RN, Tsoi KKF, Ho JMW, Lo CM, Chan FCH, Kyaw MH, et al. Global Increasing Incidence of Young-Onset Colorectal Cancer Across 5 Continents: A Joinpoint Regression Analysis of 1,922,167 Cases. Cancer Epidemiol Biomarkers Prev. 2019 Aug;28(8):1275\u0026ndash;82.\u003c/li\u003e\n\u003cli\u003eVuik FE, Nieuwenburg SA, Bardou M, Lansdorp-Vogelaar I, Dinis-Ribeiro M, Bento MJ, et al. Increasing incidence of colorectal cancer in young adults in Europe over the last 25 years. Gut. 2019 Oct;68(10):1820\u0026ndash;6.\u003c/li\u003e\n\u003cli\u003eBailey CE, Hu C-Y, You YN, Bednarski BK, Rodriguez-Bigas MA, Skibber JM, et al. Increasing disparities in the age-related incidences of colon and rectal cancers in the United States, 1975-2010. JAMA Surg. 2015 Jan;150(1):17\u0026ndash;22.\u003c/li\u003e\n\u003cli\u003eAkimoto N, Ugai T, Zhong R, Hamada T, Fujiyoshi K, Giannakis M, et al. Rising incidence of early-onset colorectal cancer - a call to action. Nat Rev Clin Oncol. 2021 Apr;18(4):230\u0026ndash;43.\u003c/li\u003e\n\u003cli\u003eZaborowski AM, Murphy B, Creavin B, Rogers AC, Kennelly R, Hanly A, et al. Clinicopathological features and oncological outcomes of patients with young-onset rectal cancer. Br J Surg. 2020 Apr;107(5):606\u0026ndash;12.\u003c/li\u003e\n\u003cli\u003eFoppa C, Tamburello S, Maroli A, Carvello M, Poliani L, Laghi L, et al. Early age of onset is an independent predictor for worse disease-free survival in sporadic rectal cancer patients. A comparative analysis of 980 consecutive patients. Eur J Surg Oncol. 2022 Apr;48(4):857\u0026ndash;63.\u003c/li\u003e\n\u003cli\u003eMcClelland PH-T, Liu T, Ozuner G. Early-Onset Colorectal Cancer in Patients under 50 Years of Age: Demographics, Disease Characteristics, and Survival. Clin Colorectal Cancer. 2022 Jun;21(2):e135\u0026ndash;44.\u003c/li\u003e\n\u003cli\u003eLieu CH, Golemis EA, Serebriiskii IG, Newberg J, Hemmerich A, Connelly C, et al. Comprehensive genomic landscapes in early and later onset colorectal cancer. Clin Cancer Res. 2019 Oct 1;25(19):5852\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eUgai T, Haruki K, Harrison TA, Cao Y, Qu C, Chan AT, et al. Molecular Characteristics of Early-Onset Colorectal Cancer According to Detailed Anatomical Locations: Comparison With Later-Onset Cases. Am J Gastroenterol. 2023 Apr 1;118(4):712\u0026ndash;26.\u003c/li\u003e\n\u003cli\u003eBerg M, Agesen TH, Thiis-Evensen E, INFAC-study group, Merok MA, Teixeira MR, et al. Distinct high resolution genome profiles of early onset and late onset colorectal cancer integrated with gene expression data identify candidate susceptibility loci. Mol Cancer. 2010 May 6;9:100.\u003c/li\u003e\n\u003cli\u003eWillauer AN, Liu Y, Pereira AAL, Lam M, Morris JS, Raghav KPS, et al. Clinical and molecular characterization of early-onset colorectal cancer. Cancer. 2019 Jun 15;125(12):2002\u0026ndash;10.\u003c/li\u003e\n\u003cli\u003ePerea J, Rueda D, Canal A, Rodr\u0026iacute;guez Y, \u0026Aacute;lvaro E, Osorio I, et al. Age at onset should be a major criterion for subclassification of colorectal cancer. J Mol Diagn. 2014 Jan;16(1):116\u0026ndash;26.\u003c/li\u003e\n\u003cli\u003eAntelo M, Balaguer F, Shia J, Shen Y, Hur K, Moreira L, et al. A high degree of LINE-1 hypomethylation is a unique feature of early-onset colorectal cancer. PLoS ONE. 2012 Sep 25;7(9):e45357.\u003c/li\u003e\n\u003cli\u003eBaba Y, Huttenhower C, Nosho K, Tanaka N, Shima K, Hazra A, et al. Epigenomic diversity of colorectal cancer indicated by LINE-1 methylation in a database of 869 tumors. Mol Cancer. 2010 May 27;9:125.\u003c/li\u003e\n\u003cli\u003eAkimoto N, Zhao M, Ugai T, Zhong R, Lau MC, Fujiyoshi K, et al. Tumor Long Interspersed Nucleotide Element-1 (LINE-1) Hypomethylation in Relation to Age of Colorectal Cancer Diagnosis and Prognosis. Cancers (Basel). 2021 Apr 22;13(9).\u003c/li\u003e\n\u003cli\u003eGogna P, O\u0026rsquo;Sullivan DE, King WD. The effect of inflammation-related lifestyle exposures and interactions with gene variants on long interspersed nuclear element-1 DNA methylation. Epigenomics. 2018 Jun 11;10(6):785\u0026ndash;96.\u003c/li\u003e\n\u003cli\u003eMartin EM, Fry RC. Environmental influences on the epigenome: Exposure-associated DNA methylation in human populations. Annu Rev Public Health. 2018 Apr 1;39:309\u0026ndash;33.\u003c/li\u003e\n\u003cli\u003eCho YH, Jang Y, Woo HD, Kim YJ, Kim SY, Christensen S, et al. LINE-1 hypomethylation is associated with radiation-induced genomic instability in industrial radiographers. Environ Mol Mutagen. 2019 Mar;60(2):174\u0026ndash;84.\u003c/li\u003e\n\u003cli\u003eMork ME, You YN, Ying J, Bannon SA, Lynch PM, Rodriguez-Bigas MA, et al. High prevalence of hereditary cancer syndromes in adolescents and young adults with colorectal cancer. J Clin Oncol. 2015 Nov 1;33(31):3544\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eStoffel EM, Koeppe E, Everett J, Ulintz P, Kiel M, Osborne J, et al. Germline genetic features of young individuals with colorectal cancer. Gastroenterology. 2018 Mar;154(4):897-905.e1.\u003c/li\u003e\n\u003cli\u003eSyed AR, Thakkar P, Horne ZD, Abdul-Baki H, Kochhar G, Farah K, et al. Old vs new: Risk factors predicting early onset colorectal cancer. World J Gastrointest Oncol. 2019 Nov 15;11(11):1011\u0026ndash;20.\u003c/li\u003e\n\u003cli\u003eConnell LC, Mota JM, Braghiroli MI, Hoff PM. The rising incidence of younger patients with colorectal cancer: questions about screening, biology, and treatment. Curr Treat Options Oncol. 2017 Apr;18(4):23.\u003c/li\u003e\n\u003cli\u003ePleguezuelos-Manzano C, Puschhof J, Rosendahl Huber A, van Hoeck A, Wood HM, Nomburg J, et al. Mutational signature in colorectal cancer caused by genotoxic pks+ E. coli. Nature. 2020 Apr;580(7802):269\u0026ndash;73.\u003c/li\u003e\n\u003cli\u003eRosendahl Huber A, Pleguezuelos-Manzano C, Puschhof J, Ubels J, Boot C, Saftien A, et al. Improved detection of colibactin-induced mutations by genotoxic E. coli in organoids and colorectal cancer. Cancer Cell. 2024 Mar 11;42(3):487-496.e6.\u003c/li\u003e\n\u003cli\u003eMa J, Rebholz CM, Braun KVE, Reynolds LM, Aslibekyan S, Xia R, et al. Whole Blood DNA Methylation Signatures of Diet Are Associated With Cardiovascular Disease Risk Factors and All-Cause Mortality. Circ Genom Precis Med. 2020 Aug;13(4):e002766.\u003c/li\u003e\n\u003cli\u003eLiu C, Marioni RE, Hedman \u0026Aring;K, Pfeiffer L, Tsai PC, Reynolds LM, et al. A DNA methylation biomarker of alcohol consumption. Mol Psychiatry. 2018 Feb;23(2):422\u0026ndash;33.\u003c/li\u003e\n\u003cli\u003eMadden RA, McCartney DL, Walker RM, Hillary RF, Bermingham ML, Rawlik K, et al. Birth weight associations with DNA methylation differences in an adult population. Epigenetics. 2021;16(7):783\u0026ndash;96.\u003c/li\u003e\n\u003cli\u003eDo WL, Sun D, Meeks K, Dugu\u0026eacute; P-A, Demerath E, Guan W, et al. Epigenome-wide meta-analysis of BMI in nine cohorts: Examining the utility of epigenetically predicted BMI. Am J Hum Genet. 2023 Feb 2;110(2):273\u0026ndash;83.\u003c/li\u003e\n\u003cli\u003eKarabegović I, Portilla-Fernandez E, Li Y, Ma J, Maas SCE, Sun D, et al. Epigenome-wide association meta-analysis of DNA methylation with coffee and tea consumption. Nat Commun. 2021 May 14;12(1):2830.\u003c/li\u003e\n\u003cli\u003evan Dongen J, Bonder MJ, Dekkers KF, Nivard MG, van Iterson M, Willemsen G, et al. DNA methylation signatures of educational attainment. NPJ Sci Learn. 2018 Mar 23;3:7.\u003c/li\u003e\n\u003cli\u003eKoh I-U, Choi N-H, Lee K, Yu H-Y, Yun JH, Kong J-H, et al. Obesity susceptible novel DNA methylation marker on regulatory region of inflammation gene: results from the Korea Epigenome Study (KES). BMJ Open Diabetes Res Care. 2020 Aug;8(1).\u003c/li\u003e\n\u003cli\u003eJoehanes R, Just AC, Marioni RE, Pilling LC, Reynolds LM, Mandaviya PR, et al. Epigenetic signatures of cigarette smoking. Circ Cardiovasc Genet. 2016 Oct;9(5):436\u0026ndash;47.\u003c/li\u003e\n\u003cli\u003eMaas SCE, Vidaki A, Wilson R, Teumer A, Liu F, van Meurs JBJ, et al. Validated inference of smoking habits from blood with a finite DNA methylation marker set. Eur J Epidemiol. 2019 Nov;34(11):1055\u0026ndash;74.\u003c/li\u003e\n\u003cli\u003ede F C Lichtenfels AJ, van der Plaat DA, de Jong K, van Diemen CC, Postma DS, Nedeljkovic I, et al. Long-term Air Pollution Exposure, Genome-wide DNA Methylation and Lung Function in the LifeLines Cohort Study. Environ Health Perspect. 2018 Feb 6;126(2):027004.\u003c/li\u003e\n\u003cli\u003ePlusquin M, Guida F, Polidoro S, Vermeulen R, Raaschou-Nielsen O, Campanella G, et al. DNA methylation and exposure to ambient air pollution in two prospective cohorts. Environ Int. 2017 Nov;108:127\u0026ndash;36.\u003c/li\u003e\n\u003cli\u003eSayols-Baixeras S, Fern\u0026aacute;ndez-Sanl\u0026eacute;s A, Prats-Uribe A, Subirana I, Plusquin M, K\u0026uuml;nzli N, et al. Association between long-term air pollution exposure and DNA methylation: The REGICOR study. Environ Res. 2019 Sep;176:108550.\u003c/li\u003e\n\u003cli\u003eGondalia R, Baldassari A, Holliday KM, Justice AE, M\u0026eacute;ndez-Gir\u0026aacute;ldez R, Stewart JD, et al. Methylome-wide association study provides evidence of particulate matter air pollution-associated DNA methylation. Environ Int. 2019 Nov;132:104723.\u003c/li\u003e\n\u003cli\u003ePanni T, Mehta AJ, Schwartz JD, Baccarelli AA, Just AC, Wolf K, et al. Genome-Wide Analysis of DNA Methylation and Fine Particulate Matter Air Pollution in Three Study Populations: KORA F3, KORA F4, and the Normative Aging Study. Environ Health Perspect. 2016 Jul;124(7):983\u0026ndash;90.\u003c/li\u003e\n\u003cli\u003eCurtis SW, Cobb DO, Kilaru V, Terrell ML, Marder ME, Barr DB, et al. Genome-wide DNA methylation differences and polychlorinated biphenyl (PCB) exposure in a US population. Epigenetics. 2021 Mar;16(3):338\u0026ndash;52.\u003c/li\u003e\n\u003cli\u003eHoang TT, Qi C, Paul KC, Lee M, White JD, Richards M, et al. Epigenome-Wide DNA Methylation and Pesticide Use in the Agricultural Lung Health Study. Environ Health Perspect. 2021 Sep 13;129(9):97008.\u003c/li\u003e\n\u003cli\u003eCancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature. 2012 Jul 18;487(7407):330\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eD\u0026iacute;ez-Villanueva A, Sanz-Pamplona R, Carreras-Torres R, Moratalla-Navarro F, Alonso MH, Par\u0026eacute;-Brunet L, et al. DNA methylation events in transcription factors and gene expression changes in colon cancer. Epigenomics. 2020 Sep 22;12(18):1593\u0026ndash;610.\u003c/li\u003e\n\u003cli\u003eNaumov VA, Generozov EV, Zaharjevskaya NB, Matushkina DS, Larin AK, Chernyshov SV, et al. Genome-scale analysis of DNA methylation in colorectal cancer using Infinium HumanMethylation450 BeadChips. Epigenetics. 2013 Sep;8(9):921\u0026ndash;34.\u003c/li\u003e\n\u003cli\u003eHa YJ, Kim CW, Roh SA, Cho DH, Park JL, Kim SY, et al. Epigenetic regulation of KLHL34 predictive of pathologic response to preoperative chemoradiation therapy in rectal cancer patients. Int J Radiat Oncol Biol Phys. 2015 Mar 1;91(3):650\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eBarrow TM, Klett H, Toth R, B\u0026ouml;hm J, Gigic B, Habermann N, et al. Smoking is associated with hypermethylation of the APC 1A promoter in colorectal cancer: the ColoCare Study. J Pathol. 2017 Nov;243(3):366\u0026ndash;75.\u003c/li\u003e\n\u003cli\u003eQu X, Sandmann T, Frierson H, Fu L, Fuentes E, Walter K, et al. Integrated genomic analysis of colorectal cancer progression reveals activation of EGFR through demethylation of the EREG promoter. Oncogene. 2016 Dec 15;35(50):6403\u0026ndash;15.\u003c/li\u003e\n\u003cli\u003eHorvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14(10):R115.\u003c/li\u003e\n\u003cli\u003eRamai D, Barakat M, Dhaliwal A, Dhindsa B, Chandan S, Ofosu A, et al. Gender and racial disparities in colorectal cancer incidence and mortality: a national cancer registry study. Int J Colorectal Dis. 2021 Aug;36(8):1801\u0026ndash;4.\u003c/li\u003e\n\u003cli\u003eZhong Y, Zhu Y, Li Q, Wang F, Ge X, Zhou G, et al. Association between Mediterranean diet adherence and colorectal cancer: a dose-response meta-analysis. Am J Clin Nutr. 2020 Jun 1;111(6):1214\u0026ndash;25.\u003c/li\u003e\n\u003cli\u003eDoubeni CA, Laiyemo AO, Major JM, Schootman M, Lian M, Park Y, et al. Socioeconomic status and the risk of colorectal cancer: an analysis of more than a half million adults in the National Institutes of Health-AARP Diet and Health Study. Cancer. 2012 Jul 15;118(14):3636\u0026ndash;44.\u003c/li\u003e\n\u003cli\u003eBotteri E, Borroni E, Sloan EK, Bagnardi V, Bosetti C, Peveri G, et al. Smoking and Colorectal Cancer Risk, Overall and by Molecular Subtypes: A Meta-Analysis. Am J Gastroenterol. 2020 Dec;115(12):1940\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eMandic M, Safizadeh F, Niedermaier T, Hoffmeister M, Brenner H. Association of overweight, obesity, and recent weight loss with colorectal cancer risk. JAMA Netw Open. 2023 Apr 3;6(4):e239556.\u003c/li\u003e\n\u003cli\u003eAlexandrov LB, Jones PH, Wedge DC, Sale JE, Campbell PJ, Nik-Zainal S, et al. Clock-like mutational processes in human somatic cells. Nat Genet. 2015 Dec;47(12):1402\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eAlexandrov LB, Kim J, Haradhvala NJ, Huang MN, Tian Ng AW, Wu Y, et al. The repertoire of mutational signatures in human cancer. Nature. 2020 Feb 5;578(7793):94\u0026ndash;101.\u003c/li\u003e\n\u003cli\u003eSia EA, Kokoska RJ, Dominska M, Greenwell P, Petes TD. Microsatellite instability in yeast: dependence on repeat unit size and DNA mismatch repair genes. Mol Cell Biol. 1997 May;17(5):2851\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eBoland CR, Goel A. Microsatellite instability in colorectal cancer. Gastroenterology. 2010 Jun;138(6):2073-2087.e3.\u003c/li\u003e\n\u003cli\u003eArnold M, Abnet CC, Neale RE, Vignat J, Giovannucci EL, McGlynn KA, et al. Global burden of 5 major types of gastrointestinal cancer. Gastroenterology. 2020 Jul;159(1):335-349.e15.\u003c/li\u003e\n\u003cli\u003eHuyghe JR, Bien SA, Harrison TA, Kang HM, Chen S, Schmit SL, et al. Discovery of common and rare genetic risk variants for colorectal cancer. Nat Genet. 2019 Jan;51(1):76\u0026ndash;87.\u003c/li\u003e\n\u003cli\u003eSteele SR, Park GE, Johnson EK, Martin MJ, Stojadinovic A, Maykel JA, et al. The impact of age on colorectal cancer incidence, treatment, and outcomes in an equal-access health care system. Dis Colon Rectum. 2014 Mar;57(3):303\u0026ndash;10.\u003c/li\u003e\n\u003cli\u003eChen N, Liang H, Huang T, Huang N. Exposome approach for identifying modifiable factors for the prevention of colorectal cancer. Sci Rep. 2022 Dec 14;12(1):21615.\u003c/li\u003e\n\u003cli\u003eCarr PR, Weigl K, Edelmann D, Jansen L, Chang-Claude J, Brenner H, et al. Estimation of Absolute Risk of Colorectal Cancer Based on Healthy Lifestyle, Genetic Risk, and Colonoscopy Status in a Population-Based Study. Gastroenterology. 2020 Jul;159(1):129-138.e9.\u003c/li\u003e\n\u003cli\u003eHuang P, Liu Y. A reasonable diet promotes balance of intestinal microbiota: prevention of precolorectal cancer. Biomed Res Int. 2019 Jul 25;2019:3405278.\u003c/li\u003e\n\u003cli\u003eWilkins T, McMechan D, Talukder A. Colorectal cancer screening and prevention. Am Fam Physician. 2018 May 15;97(10):658\u0026ndash;65.\u003c/li\u003e\n\u003cli\u003eBrooke HL, Talb\u0026auml;ck M, Martling A, Feychting M, Ljung R. Socioeconomic position and incidence of colorectal cancer in the Swedish population. Cancer Epidemiol. 2016 Feb;40:188\u0026ndash;95.\u003c/li\u003e\n\u003cli\u003eEgeberg R, Halkjaer J, Rottmann N, Hansen L, Holten I. Social inequality and incidence of and survival from cancers of the colon and rectum in a population-based study in Denmark, 1994-2003. Eur J Cancer. 2008 Sep;44(14):1978\u0026ndash;88.\u003c/li\u003e\n\u003cli\u003eSarma EA, Kawachi I, Poole EM, Tworoger SS, Giovannucci EL, Fuchs CS, et al. Social integration and survival after diagnosis of colorectal cancer. Cancer. 2018 Feb 15;124(4):833\u0026ndash;40.\u003c/li\u003e\n\u003cli\u003eGarc\u0026iacute;a-P\u0026eacute;rez J, Fern\u0026aacute;ndez de Larrea-Baz N, Lope V, Molina AJ, O\u0026rsquo;Callaghan-Gordo C, Alonso MH, et al. Residential proximity to industrial pollution sources and colorectal cancer risk: A multicase-control study (MCC-Spain). Environ Int. 2020 Nov;144:106055.\u003c/li\u003e\n\u003cli\u003eJenwitheesuk K, Peansukwech U, Jenwitheesuk K. Accumulated ambient air pollution and colon cancer incidence in Thailand. Sci Rep. 2020 Oct 20;10(1):17765.\u003c/li\u003e\n\u003cli\u003ePritchett N, Spangler EC, Gray GM, Livinski AA, Sampson JN, Dawsey SM, et al. Exposure to Outdoor Particulate Matter Air Pollution and Risk of Gastrointestinal Cancers in Adults: A Systematic Review and Meta-Analysis of Epidemiologic Evidence. Environ Health Perspect. 2022 Mar 2;130(3):36001.\u003c/li\u003e\n\u003cli\u003eL\u0026oacute;pez-Abente G, Garc\u0026iacute;a-P\u0026eacute;rez J, Fern\u0026aacute;ndez-Navarro P, Boldo E, Ramis R. Colorectal cancer mortality and industrial pollution in Spain. BMC Public Health. 2012 Aug 1;12:589.\u003c/li\u003e\n\u003cli\u003eWei Y, Danesh Yazdi M, Ma T, Castro E, Liu CS, Qiu X, et al. Additive effects of 10-year exposures to PM2.5 and NO2 and primary cancer incidence in American older adults. Environ Epidemiol. 2023 Aug 1;7(4):e265.\u003c/li\u003e\n\u003cli\u003eFeng J, Cavallero S, Hsiai T, Li R. Impact of air pollution on intestinal redox lipidome and microbiome. Free Radic Biol Med. 2020 May 1;151:99\u0026ndash;110.\u003c/li\u003e\n\u003cli\u003eJin Y, Lu L, Tu W, Luo T, Fu Z. Impacts of polystyrene microplastic on the gut barrier, microbiota and metabolism of mice. Sci Total Environ. 2019 Feb 1;649:308\u0026ndash;17.\u003c/li\u003e\n\u003cli\u003eXia T, Guo J, Zhang B, Song C, Zhao Q, Cui B, et al. Bisphenol A Promotes the Progression of Colon Cancer Through Dual-Targeting of NADPH Oxidase and Mitochondrial Electron-Transport Chain to Produce ROS and Activating HIF-1\u0026alpha;/VEGF/PI3K/AKT Axis. Front Endocrinol (Lausanne). 2022 Jul 4;13:933051.\u003c/li\u003e\n\u003cli\u003eUnited States Environmental Protection Agency. US EPA - Pesticides - Fact Sheet for Picloram. 1995 Aug [cited 2024 May 13]; Available from: https://www3.epa.gov/pesticides/chem_search/reg_actions/reregistration/fs_PC-005101_1-Aug-95.pdf\u003c/li\u003e\n\u003cli\u003eDavoren MJ, Schiestl RH. Glyphosate-based herbicides and cancer risk: a post-IARC decision review of potential mechanisms, policy and avenues of research. Carcinogenesis. 2018 Oct 8;39(10):1207\u0026ndash;15.\u003c/li\u003e\n\u003cli\u003eSass JB, Colangelo A. European Union bans atrazine, while the United States negotiates continued use. Int J Occup Environ Health. 2006 Sep;12(3):260\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eTurner MC, Cogliano V, Guyton K, Madia F, Straif K, Ward EM, et al. Research Recommendations for Selected IARC-Classified Agents: Impact and Lessons Learned. Environ Health Perspect. 2023 Oct 30;131(10):105001.\u003c/li\u003e\n\u003cli\u003eBerrington de Gonz\u0026aacute;lez A, Masten SA, Bhatti P, Fortner RT, Peters S, Santonen T, et al. Advisory Group recommendations on priorities for the IARC Monographs. Lancet Oncol. 2024 May;25(5):546\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eRemigio RV, Andreotti G, Sandler DP, Erickson PA, Koutros S, Albert PS, et al. An Updated Evaluation of Atrazine-Cancer Incidence Associations among Pesticide Applicators in the Agricultural Health Study Cohort. Environ Health Perspect. 2024 Feb 21;132(2):27010.\u003c/li\u003e\n\u003cli\u003eO\u0026rsquo;Sullivan DE, Sutherland RL, Town S, Chow K, Fan J, Forbes N, et al. Risk Factors for Early-Onset Colorectal Cancer: A Systematic Review and Meta-analysis. Clin Gastroenterol Hepatol. 2022 Jun;20(6):1229-1240.e5.\u003c/li\u003e\n\u003cli\u003eCercek A, Chatila WK, Yaeger R, Walch H, Fernandes GDS, Krishnan A, et al. A Comprehensive Comparison of Early-Onset and Average-Onset Colorectal Cancers. J Natl Cancer Inst. 2021 Nov 29;113(12):1683\u0026ndash;92.\u003c/li\u003e\n\u003cli\u003eSong N, Shin A, Park JW, Kim J, Oh JH. Common risk variants for colorectal cancer: an evaluation of associations with age at cancer onset. Sci Rep. 2017 Jan 13;7:40644.\u003c/li\u003e\n\u003cli\u003eChen FW, Sundaram V, Chew TA, Ladabaum U. Advanced-Stage Colorectal Cancer in Persons Younger Than 50 Years Not Associated With Longer Duration of Symptoms or Time to Diagnosis. Clin Gastroenterol Hepatol. 2017 May;15(5):728-737.e3.\u003c/li\u003e\n\u003cli\u003eSilva TC, Colaprico A, Olsen C, D\u0026rsquo;Angelo F, Bontempi G, Ceccarelli M, et al. TCGA Workflow: Analyze cancer genomics and epigenomics data using Bioconductor packages. [version 2; peer review: 1 approved, 2 approved with reservations]. F1000Res. 2016 Jun 29;5:1542.\u003c/li\u003e\n\u003cli\u003eColaprico A, Silva TC, Olsen C, Garofano L, Cava C, Garolini D, et al. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res. 2016 May 5;44(8):e71.\u003c/li\u003e\n\u003cli\u003eMounir M, Lucchetta M, Silva TC, Olsen C, Bontempi G, Chen X, et al. New functionalities in the TCGAbiolinks package for the study and integration of cancer data from GDC and GTEx. PLoS Comput Biol. 2019 Mar 5;15(3):e1006701.\u003c/li\u003e\n\u003cli\u003eDavis S, Meltzer PS. GEOquery: a bridge between the Gene Expression Omnibus (GEO) and BioConductor. Bioinformatics. 2007 Jul 15;23(14):1846\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eDing W, Kaur D, Horvath S, Zhou W. Comparative epigenome analysis using Infinium DNA methylation BeadChips. Brief Bioinformatics. 2023 Jan 19;24(1).\u003c/li\u003e\n\u003cli\u003eZhou W, Hinoue T, Barnes B, Mitchell O, Iqbal W, Lee SM, et al. DNA methylation dynamics and dysregulation delineated by high-throughput profiling in the mouse. Cell Genomics. 2022 Jul 13;2(7).\u003c/li\u003e\n\u003cli\u003eZhou W, Triche TJ, Laird PW, Shen H. SeSAMe: reducing artifactual detection of DNA methylation by Infinium BeadChips in genomic deletions. Nucleic Acids Res. 2018 Nov 16;46(20):e123.\u003c/li\u003e\n\u003cli\u003eTriche TJ, Weisenberger DJ, Van Den Berg D, Laird PW, Siegmund KD. Low-level processing of Illumina Infinium DNA Methylation BeadArrays. Nucleic Acids Res. 2013 Apr;41(7):e90.\u003c/li\u003e\n\u003cli\u003eHastie T, Tibshirani R, Narasimhan B, Chu G. Impute: Imputation for microarray data. 2016 Jan 1;17:520\u0026ndash;5.\u003c/li\u003e\n\u003cli\u003eDu P, Kibbe WA, Lin SM. lumi: a pipeline for processing Illumina microarray. Bioinformatics. 2008 Jul 1;24(13):1547\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eDu P, Zhang X, Huang C-C, Jafari N, Kibbe WA, Hou L, et al. Comparison of Beta-value and M-value methods for quantifying methylation levels by microarray analysis. BMC Bioinformatics. 2010 Nov 30;11:587.\u003c/li\u003e\n\u003cli\u003eLin SM, Du P, Huber W, Kibbe WA. Model-based variance-stabilizing transformation for Illumina microarray data. Nucleic Acids Res. 2008 Feb;36(2):e11.\u003c/li\u003e\n\u003cli\u003e\u0026emsp;Du P, Kibbe WA, Lin SM. nuID: a universal naming scheme of oligonucleotides for illumina, affymetrix, and other microarrays. Biol Direct. 2007 May 31;2:16.\u003c/li\u003e\n\u003cli\u003e\u0026emsp;Chen Y, Lemire M, Choufani S, Butcher DT, Grafodatskaya D, Zanke BW, et al. Discovery of cross-reactive probes and polymorphic CpGs in the Illumina Infinium HumanMethylation450 microarray. Epigenetics. 2013 Feb;8(2):203\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003e\u0026emsp;Viechtbauer W. Conducting Meta-Analyses in R with the metafor Package. J Stat Softw. 2010;36(3).\u003c/li\u003e\n\u003cli\u003e\u0026emsp;Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society: Series B (Methodological). 1995 Jan;57(1):289\u0026ndash;300.\u003c/li\u003e\n\u003cli\u003e\u0026emsp;Surveillance Research Program, National Cancer Institute. Surveillance Research Program, National Cancer Institute SEER*Stat software (seer.cancer.gov/seerstat) version 8.4.1.\u003c/li\u003e\n\u003cli\u003e\u0026emsp;Kuznetsova A, Brockhoff PB, Christensen RHB. lmertest package: tests in linear mixed effects models. J Stat Softw. 2017;82(13):1\u0026ndash;26.\u003c/li\u003e\n\u003cli\u003e\u0026emsp;Team RRC. R: A language and environment for statistical computing. 2013;\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4528579/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4528579/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe incidence of colorectal cancer (CRC) is rapidly rising in individuals younger than 50, particularly in high-income countries. This rise parallels shifts in lifestyle and environmental factors, collectively termed the exposome; however, whether these are causally linked to the development of early-onset CRC (EOCRC) has not been investigated. Due to limited exposome data in most cancer cohorts, we constructed weighted methylation risk scores (MRS) as proxies for exposome exposure to pinpoint specific risk factors associated with EOCRC. Our analysis confirms previously identified risk factors, such as educational attainment, diet, and smoking habits. Moreover, we identified the exposure to the herbicide picloram as a novel risk factor (Padj. = 0.00049), a result we replicated in a meta-analysis comprising six CRC cohorts (P = 0.021), comparing EOCRC cases with patients diagnosed aged ≥70. Subsequently, we employed population-based data from 81 U.S. counties over 20 years and validated the association between picloram usage and EOCRC incidence (P = 2.87×10\u003csup\u003e-3\u003c/sup\u003e). These findings highlight the critical role of the exposome in EOCRC risk, underscoring the urgency for targeted personal and policy-level interventions.\u003c/p\u003e","manuscriptTitle":"Epigenetic Fingerprints Link Early-Onset Colon and Rectal Cancer to Pesticide Exposure","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-12 17:09:21","doi":"10.21203/rs.3.rs-4528579/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-medicine","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nm","sideBox":"Learn more about [Nature Medicine](http://www.nature.com/nm/)","snPcode":"","submissionUrl":"","title":"Nature Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"c0bf1d8d-1d8f-442c-8132-172bf1324f8c","owner":[],"postedDate":"June 12th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":33087480,"name":"Biological sciences/Cancer/Gastrointestinal cancer/Colorectal cancer"},{"id":33087481,"name":"Biological sciences/Computational biology and bioinformatics/Data integration"},{"id":33087482,"name":"Health sciences/Risk factors"},{"id":33087483,"name":"Biological sciences/Molecular biology/Epigenetics/DNA methylation"},{"id":33087484,"name":"Biological sciences/Cancer/Cancer epidemiology"}],"tags":[],"updatedAt":"2026-04-23T07:15:06+00:00","versionOfRecord":{"articleIdentity":"rs-4528579","link":"https://doi.org/10.1038/s41591-026-04342-5","journal":{"identity":"nature-medicine","isVorOnly":false,"title":"Nature Medicine"},"publishedOn":"2026-04-21 04:00:00","publishedOnDateReadable":"April 21st, 2026"},"versionCreatedAt":"2024-06-12 17:09:21","video":"","vorDoi":"10.1038/s41591-026-04342-5","vorDoiUrl":"https://doi.org/10.1038/s41591-026-04342-5","workflowStages":[]},"version":"v1","identity":"rs-4528579","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4528579","identity":"rs-4528579","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}