Both overall and abdominal obesity are well-established risk factors for various cancer types, including colorectal cancer (CRC) [1]. However, how adiposity impacts CRC development has been insufficiently investigated. Three primary hypotheses have been suggested to elucidate the biological pathways that link adiposity and CRC: alterations in insulin signaling, dysregulation of adipose tissue-derived inflammation, and sex hormone metabolism [2, 3]. New mechanisms are also emerging, including altered gut microbiome and gut hormones, such as Ghrelin and nonalcoholic fatty liver disease (NAFLD). One of the key mechanisms proposed, and a common feature in most pathways, is inflammation [3]. Adiposity is associated with a systemic subclinical inflammation and higher levels of inflammatory biomarkers such as C-reactive protein (CRP), tumor necrosis factor (TNF), interleukin‑1β (IL‑1β), IL‑6, and IL‑18 [2]. Inflammation can contribute to cancer development through mechanisms, such as the production of free radicals, including reactive oxygen intermediates, by suppressing the immune system, causing abnormal cell signaling, which promotes proliferative and anti-apoptotic pathways, angiogenesis, and cell migration [3]. To quantify how much of the association between adiposity and CRC risk might be explained by inflammation, as reflected in increased serum levels of CRP —a nonspecific marker of systemic inflammation, we used body mass index (BMI) as a measure of general obesity, and waist circumference (WC) and waist-to-hip ratio (WHR) as measures of abdominal obesity, and we paid particular attention to a potential role of reverse causation due to cancer-related changes in body weight and CRP levels. Data from 499,926 UK Biobank study participants aged 40-69, collected from 22 UK Biobank assessment centers, were utilized. Detailed information on the study population and design, exposure and outcome assessments, and statistical analysis is provided in the Supplementary Materials and Methods. After excluding participants with previous cancer diagnosis (except non-melanoma skin cancer), missing BMI, WHR, WC, and CRP, 429,073 participants remained and were included in the analysis (Supplementary Figure S1). Of these, 5,544 were diagnosed with CRC during a median follow-up of 11.8 years (interquartile range: 11.0-12.5). Main characteristics of the cohort are shown in Supplementary Table S1. Median age at baseline was 57 years, 53.2% of participants were female, and 94.6% were white. Median BMI, WC, and WHR for the whole cohort population were 26.7 kg/m2, 90.0 cm, and 0.87, respectively. Furthermore, approximately 22% of the population had CRP levels greater than or equal to 3 mg/L at baseline. Individuals classified as overweight or obese exhibited elevated CRP levels compared to those with a normal BMI. Additionally, participants in higher quartiles for both WC and WHR demonstrated significantly higher CRP values compared to those in the lowest quartile (Supplementary Figure S2). Furthermore, higher CRP levels were observed across all categories of all anthropometric measures in CRC cases diagnosed within the first four years of follow-up compared to those diagnosed later, suggesting a potential influence of preclinical cancer on CRP concentrations (Supplementary Figure S3). The Spearman rank correlation coefficients for the relationship between various anthropometric measures and CRP levels was highest for BMI (0.44), followed by WC (0.38) and lowest for WHR (0.23), with stronger correlations observed for BMI and WC among women compared to men. (Supplementary Table S2). In a standard analysis including the entire follow-up time, the hazard ratios (HRs) and 95% confidence intervals (CIs) compared to normal BMI decreased from 1.12 (1.05-1.20) to 1.09 (1.02-1.17) for overweight and from 1.24 (1.15-1.34) to 1.17 (1.08-1.26) for obesity, after adjustment for the natural logarithm (ln) of CRP levels at baseline (mg/L). For WHR and WC, the associations for the highest versus lowest quartile decreased from 1.38 (1.27-1.49) to 1.32 (1.21-1.43) and from 1.35 (1.24-1.47) to 1.27 (1.17-1.39), respectively, after adjusting for ln (CRP), which by itself showed a clear association with increased CRC risk (Table 1). Model 2 (Full)b Model 3 (Full + CRP)c Initial four years Excluded Excluding the first four years of follow-up to minimize a potential role of reverse causality resulted in stronger HRs for the association between BMI and CRC risk, while the associations between WC, and WHR and CRC risk remained essentially unchanged. However, the attenuation of the association after including ln (CRP) in the models essentially disappeared for all measures of adiposity. For example, the HRs (95% CIs) for overweight and obesity compared to normal BMI were 1.13 (1.05-1.23) and 1.30 (1.19-1.42), respectively, before adjusting for ln (CRP), and 1.13 (1.04-1.22) and 1.28 (1.16-1.40) after adjustment. A similar pattern was observed for the associations between WHR and WC with CRC risk. Furthermore, CRP was no longer associated with CRC risk after exclusion of the initial four years of follow-up (Table 1). Cancer cachexia, characterized by muscle loss with or without concurrent fat loss, is common among cancer patients, including CRC, even before diagnosis [4, 5]. Hence, CRC cases diagnosed shortly after recruitment in cohort studies might have been present at the time of enrollment leading to an underestimation of BMI in those participants and consequently a very weak and even inverse BMI-CRC association in the early follow-up years. A major role of reverse causality due to prediagnostic weight loss, leading to attenuation of the association between general adiposity, as reflected by increased BMI, and CRC risk in epidemiological studies has previously been demonstrated and was also evident in our analyses [6, 7]. As mentioned, inflammation is a hallmark of cancer and is also considered a key player in carcinogenesis, including in CRC. It appears plausible to assume that part of the association between inflammatory markers and CRC risk observed in previous studies may likewise be due to reverse causality due to inflammatory processes following rather than preceding CRC development. This hypothesis is supported by our multivariable analyses, in which associations between CRP and CRC risk were consistently seen in models including the entire follow-up, but essentially disappeared in the models excluding the initial four years of follow-up. Our findings are consistent with other studies, showing strong associations between CRP and CRC risk only during the early years of follow-up and no association when these early years of follow-up were excluded (2-5 years) [8, 9]. These results do not support the role of CRP in CRC etiology. The use of anti-inflammatory agents especially aspirin has been shown to be associated with lower CRC incidence in some studies and aspirin has been recommended for CRC chemoprevention. However, whether the use of these medications reduces CRC risk remains controversial, and the evidence is currently insufficient [10]. Our findings may help to explain the difficulties and failures of anti-inflammatory chemoprevention of CRC and underline the importance of alternative approaches to CRC prevention, such as promotion of diets rich in fruits and vegetables. In the present study, we evaluated potential mediatory effects of inflammation, as reflected in elevated serum CRP levels, in the association between measures of general and abdominal obesity and CRC risk. Large sample size, comprehensive adjustment for potential confounders, and measured (vs self-reported) anthropometric measures were among the most important strengths of our study, while consideration of anthropometric measures and a single inflammatory biomarker only at baseline, a majorly white population which limits the generalizability, and potential residual confounding were among the limitations. Despite its limitations, our analysis underlines the importance to consider potential reverse causality in the analyses of the associations between adiposity, systemic inflammation and CRC risk. The patterns observed in our analyses excluding the initial four years of follow-up do suggest that factors other than CRP-defined systemic inflammation might play a more relevant role in mediating the increased CRC risk due to adiposity. A lower than previously assumed role of systemic inflammation for CRC risk could also partly explain the challenges and shortcomings of chemoprevention efforts with anti-inflammatory drugs like aspirin. The study was conceptualized by Hermann Brenner and Fatemeh Safizadeh. Fatemeh Safizadeh conducted the data analysis. Fatemeh Safizadeh and Hermann Brenner drafted the initial manuscript. Interpretation of the data was a collective effort involving Hermann Brenner, Fatemeh Safizadeh, Marko Mandic, and Michael Hoffmeister. A comprehensive revision of the manuscript was carried out with significant contributions from all authors. The finalized version of the manuscript received approval from all authors for publication. We thank UK Biobank participants. Authors have no conflict of interests to disclose. UK Biobank was established by the Wellcome Trust medical charity, Medical Research Council, Department of Health, Scottish Government and the Northwest Regional Development Agency. It has also had funding from British Heart Foundation, Cancer Research UK, Diabetes UK, and National Institute for Health Research (NIHR). UK Biobank is supported by the National Health Service (NHS). The UK Biobank was approved by the North West Multi center Research Ethics Committee (MREC) as a Research Tissue Bank (RTB) approval (renewed approval in 2021:21/NW/0157). Electronic informed consent was obtained from all individual participants included in the UK Biobank. Data was re-used with the permission of the UK Biobank. The UK Biobank is an open-access resource and bona fide researchers can apply to use the UK Biobank dataset by registering and applying at https://www.ukbiobank.ac.uk/enableyourresearch/apply-for-access. The data and analysis codes used for this study are going to be available on the UK Biobank website for registered researchers at the UK Biobank and an application fee. This research has been conducted using the UK Biobank Resource under application No 66591. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Supplementary Figure S4 shows how DEPTH common risk regions are defined.

Supplementary Table S7 shows the number of risk regions that are common to both CCFR and GECCO datasets in stage 1 DEPTH analysis, as well as whether they are detected by logistic regression analysis or were in previous CRC GWAS studies.

Supplementary Figure S6 shows the calibration on relative risk of PRS stratified by PRS with 7 bins in groups of different ancestry in the GERA cohort.

This file includes the expression imputation statistics and included SNPs from the elastin net models.

This file details the two-step interaction tests, and the gene-based aggregate test.

Studies aimed to evaluate the expected impact of alternative screening strategies are essential for optimizing colorectal cancer (CRC) screening offers, but such studies are lacking in Germany, where two screening colonoscopies (CS) 10 years apart are offered for men from age 50 and women from age 55. Our aim was to explore whether and to what extent the efficacy of utilizing two CS could be enhanced by alternative starting ages and screening intervals. We modeled the expected numbers of CRC cases, CRC deaths, years of potential life lost (YPLL), and disability-adjusted life years (DALYs) due to CRC in hypothetical cohorts of 100,000 men and women aged 45-85 using COSIMO, a validated Markov-based multi-state simulation model. Modeled strategies included combinations of starting ages (45/50/55/60) and CS (10/15/20 years). For men, CRC deaths could be slightly reduced by extending the interval to 15 years, with a second CS at 65. YPLL and DALYs would be reduced by decreasing starting age to 45 when combined with a 15-year screening interval. For women, use of two CS at ages 50 and 65 would reduce all CRC burden parameters compared to the current earliest-use offer at 55 and 65 years. Our results suggest that lowering the starting age of screening colonoscopy to 45 for men and 50 for women, combined with extending the CS screening interval to 15 years would have the potential to enable significant reductions in years of potential life lost, and disability-adjusted life years compared to current screening offers in Germany.

Supplementary Table S2 shows the descriptive statistics of GERA study participants by racial/ethnic groups and sex.

This figure depicts the flowchart of the two-step interaction test.

Current evidence suggests that the association between non-steroidal anti-inflammatory drug (NSAID) use and colorectal cancer (CRC) risk remains uncertain, and little is known about potential heterogeneity by molecular subtype. Clarifying these associations may provide insights into CRC carcinogenesis. The aim of this systematic review was to evaluate the association between NSAID use and CRC risk by four clinically relevant molecular subtypes. PubMed, Web of Science and Embase were searched for studies assessing NSAID use in relation CRC risk by microsatellite instability (MSI), the CpG island methylator phenotype (CIMP), somatic mutations in the B-Raf proto-oncogene serine/threonine kinase (BRAF) gene or the Kirsten rat sarcoma viral oncogene homolog gene (KRAS) status. Meta-analyses were performed to calculate summary relative risks (sRR). Nine studies met the inclusion criteria (n = 4,726 overall study population). NSAID use was associated with a stronger inverse association for BRAF-wildtype CRC (sRR = 0.73; 95% CI: 0.66–0.80) than for BRAF-mutated CRC (sRR = 0.85; 95% CI: 0.71–1.02), although the difference was not statistically significant (P heterogeneity = 0.11). Associations were in a similar direction for CIMP-high (sRR = 0.71; 95% CI: 0.61–0.83) and CIMP-low CRC (sRR = 0.73; 95% CI: 0.66–0.80), with no difference by CIMP status (P heterogeneity = 0.80). Similarly, risk reductions were observed for KRAS-mutated (sRR = 0.78; 95% CI: 0.69–0.88) and KRAS-wildtype (sRR = 0.76; 95% CI: 0.69–0.84), and for both MSI-high (sRR = 0.80; 95% CI: 0.67–0.96) and MSI-low/MSS CRC (sRR = 0.75; 95% CI: 0.69–0.82), with no evidence of heterogeneity by KRAS or MSI status (P heterogeneity = 0.76 and P heterogeneity = 0.51 respectively). NSAID use was associated with lower risk of CRC across molecular subtypes, with suggestive differences by BRAF mutation status. These findings highlight the need for further studies to clarify potential subtype-specific associations and their implications for targeted chemoprevention strategies. Mary Jose. Urruchúa-Rodríguez, Toktam Pour, Hermann Brenner, Michael Hoffmeister. Non-steroidal anti-inflammatory drugs use and risk of colorectal cancer molecular subtypes: A systematic review and meta-analysis [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: The Rise in Early-Onset Cancers—Knowledge Gaps and Research Opportunities; 2025 Dec 10-13; Montreal, QC, Canada. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(23_Suppl):Abstract nr B036.

This figure depicts the two-step EDGE test plots of prioritized SNPs.

This figure depicts the LocusZoom plots for SNPs interacting with smoking intensity for colorectal cancer risk.

Supplementary Table S4 shows the number of participants filtered out at each QC stage (according to Supplementary Table S2).

This diagnostic study compares the specificity and sensitivity of comercial fecal immunochemical tests with a multitarget stool DNA test.

Supplementary Tables 1-4 Table S1 shows SNPs previously reported to be associated with oxaliplatin treatment. Table S2 presents detectable effect sizes at 85% power and type I error of 0.05 and 4.7 x 10-4 (corresponding to Bonferroni corrected p-value) for all endpoints based on minor allele frequency of genetic variants. Table S3-4 shows SNP associations with first-line oxaliplatin treatment for all endpoints in stage II-III CRC patients (Table S3) and mCRC patients (Table S4).

Supplementary Table S7 shows the 10-year time-dependent AUC estimates of the PRS-enhanced model in the GERA European- ancestry participants.

Supplementary Table S1 shows the descriptive characteristics of study populations in GECCO and CORECT.

Supplementary Table S3 shows the comparison on characteristics between GECCO/CORECT study and GERA Europeans-ancestry participants

BACKGROUND:Excess alcohol consumption is associated with increased risk of colorectal cancer, but the evidence on the individual and joint effects of alcohol consumption and genetic risk on the occurrence of various stages of colorectal carcinogenesis is limited. METHODS:We evaluated the associations of alcohol consumption and a polygenic risk score based on 140 colorectal cancer related loci with findings of colorectal neoplasia among 4662 participants in the German screening colonoscopy program. Analyses were conducted by multiple logistic regression. We determined genetic risk equivalents to quantify the effect of alcohol consumption in terms of the difference in polygenic risk score conveying equivalent risk. RESULTS:Moderate and high (12 to <25 g/d and ≥25 g/d) alcohol consumption was associated with increased risk of advanced colorectal neoplasia (adjusted odds ratio = 1.28, 95% CI = 1.03 to 1.58, and adjusted odds ratio = 1.44, 95% CI = 1.14 to 1.81, respectively), while associations with any colorectal neoplasia were weaker. No significant interactions between alcohol consumption and polygenic risk score were observed. Participants with high alcohol consumption in the highest polygenic risk score tertile had a 3.4-fold increased risk of advanced neoplasia compared with individuals with low or no alcohol consumption in the lowest polygenic risk score tertile. The estimated impact of high alcohol consumption on the risk of advanced neoplasia was equivalent to the risk increase by a 26-percentile-higher polygenic risk score (genetic risk equivalent = 26, 95% CI = 9 to 44). CONCLUSION:High alcohol consumption and polygenic risk score have a major impact on the risk of advanced colorectal neoplasia. The estimated preventive impact of avoiding high alcohol consumption is as strong as the impact of having a substantially lower polygenic risk.

This table includes the smoking habits descriptives per studies included in the analysis.