Abstract
Background
Serum urate is the most abundant antioxidant molecule in human blood and may play a role in cancer prevention. However, the association between serum urate levels and colorectal cancer (CRC) risk remains inconclusive, with the underlying causal mechanisms still undefined.
Methods
In the UK Biobank (UKB) cohort, we investigated the prospective association between serum urate levels and the risk of CRC. We used Cox proportional hazards models to estimate the multivariable hazard ratios (HR) for CRC. Subgroup analyses were performed based on anatomical subsite and sex. Additionally, two-sample Mendelian randomization (MR) analysis was conducted to assess the potential causal effect of genetically determined urate levels on CRC risk.
Results
During a median follow-up of 11.58 years, 1960 CRC events were recorded among 180,480 participants without baseline CRC in the UKB. Higher urate levels were associated with a decreased risk of CRC (HR = 0.85, 95% CI: 0.72–0.99, P = 0.038). Moreover, the association between urate levels and CRC risk varied by anatomical subsite and sex. Higher urate levels were associated with a decreased risk of colon cancer in the overall population (HR = 0.79, 95% CI: 0.65–0.96, P = 0.015) and colon cancer in female (HR = 0.66, 95% CI: 0.50–0.87, P = 0.003), but not with rectal cancer or CRC in male. MR results supported a causal relationship between serum urate and CRC risk, with higher serum urate levels corresponding to a lower risk of CRC.
Conclusions
In this study, higher urate levels were associated with a reduced risk of CRC, and MR analysis supported a potential causal relationship between urate levels and CRC risk. However, further studies are needed to confirm causality and to investigate the potential subsite-specific effects of urate on CRC development.
Supplementary Information
The online version contains supplementary material available at 10.1007/s12672-025-03924-y.
Keywords: Urate, Colorectal cancer, Cohort study, Mendelian randomization
Introduction
Colorectal cancer (CRC) is notable for its high incidence and mortality rates worldwide, representing a substantial public health challenge [1]. The incidence of CRC has been escalating globally in recent years, a trend largely attributed to lifestyle changes, including increased adiposity [2], alcohol intake [3], and red meat diet [4]. A range of metabolic factors including elevated fasting plasma glucose levels, increased low-density lipoprotein (LDL) cholesterol, reduced bone mineral density, and renal dysfunction are associated with the development of CRC [5, 6]. While nonmodifiable risk factors for CRC are largely unchangeable, it is imperative to investigate the existence of potentially modifiable risk factors that could influence the development of the disease.
Urate, the end product of purine metabolism, is predominantly found in the blood as uric acid. It is not only endogenously synthesized by the liver but also absorbed from dietary sources rich in purines. Several epidemiological studies have demonstrated that elevated serum uric acid levels are associated with an increased risk of cancer incidence and mortality. Particularly, a prospective cohort study from Austria has demonstrated that adult males with higher uric acid levels face an increased likelihood of developing cancer [7]. Additionally, other research has indicated that patients with higher uric acid levels exhibit an increased risk of CRC compared to those with lower uric acid levels [8]. Conversely, several cohort studies have indicated that elevated uric acid levels are associated with a reduced risk of cancer incidence and mortality. For instance, a UK-based study that investigated the link between lung cancer and serum uric acid levels revealed an inverse correlation with the risk of developing lung cancer, an effect that appears to be confined to current smokers [9]. This inconsistency could potentially be due to the influence of confounding factors and the possibility of reverse causation on the observed associations. Furthermore, the causal relationship between urate levels and colorectal cancer outcomes is yet to be established.
Mendelian randomization (MR) is an epidemiological technique that uses genetic variants associated with specific exposures as instrumental variables to deduce potential causal effects on health outcomes [10]. Which eliminates the influence of confounding factors because alleles are randomly distributed during gamete formation and at the time of conception [11]. To clarify the relationship between urate and CRC and to establish a solid scientific basis for improving the effectiveness of health management strategies, we employed large-scale cohort data and the MR approach. Which may enable us to ascertain the causal link between serum urate levels and CRC incidence.
Materials and methods
Prospective cohort study
Data source and study participants
The UK Biobank (UKB) is a prospective study that enrolled over 500,000 individuals, aged 40 to 79 years, from 22 assessment centers across the United Kingdom between 2006 and 2010. During the recruitment phase, participants underwent comprehensive assessments by trained health professionals, which included demographic details, lifestyle factors, physical measurements, and other health-related parameters. Additionally, blood samples were collected from each participant for genotyping purposes. The study protocol for the UKB can be publicly accessed on the official website (https://www.ukbiobank.ac.uk/). This study used the UKB data under the application number 106,528.The datasets used in this study were available from public databases and therefore did not require additional ethical approval or informed consent.
The inclusion and exclusion criteria for the study population are shown in Figure S1. Among the 50,2356 participants in the UKB cohort, we excluded 223 participants who withdrew informed consent. To minimize the impact of population stratification on our results, we excluded participants who self-reported as non-white (n = 29,769). We further excluded participants with a history of cancer at baseline (n = 51,797, except non-melanoma skin cancer: C44) and those with inflammatory bowel disease (n = 2905). Information on prevalent diseases was obtained through self-reported and hospital inpatient records and medication (Table S1). Additionally, participants with missing urate levels (n = 26,551) and those with missing covariate data (n = 210,631) were excluded. The baseline characteristics of participants who were included and excluded because of covariate information were shown in Table S2. A total of 180,480 participants were included in the final analysis. The start date of the cohort was defined as the date participants enrolled at the study center, and the end date was the earliest date of cancer diagnosis, loss to follow-up, death, or the end of the follow-up period. For the analysis, the most recent cancer follow-up completion date for England was December 31, 2020, for Wales was December 31, 2016, and for Scotland was November 30, 2021.
Assessment of exposure, outcome, and covariates
The peripheral blood samples collected from UK Biobank participants were analyzed by the UK Biobank central laboratory within 24 h of sampling. Serum urate levels were measured by the uricase PAP method with a Beckman Coulter AU5800 analyzer (Beckman Coulter (UK), Ltd), details at https://biobank.ndph.ox.ac.uk/ukb/ukb/docs/haematology.pdf.
Cancer diagnosis data for UKB participants are sourced from the Health and Social Care Information Centre for those residing in England and Wales, and from the National Health Service Central Register for participants from Scotland. Cancer cases were classified using the 10th revision of the International Classification of Diseases (ICD-10) coding system. Colorectal cancer cases include colon cancer (ICD-10 code C18) and rectal cancer (ICD-10 code C19-20).
At baseline, demographic characteristics, lifestyle factors (such as smoking, alcohol consumption, physical activity), daily diet, and disease history were collected through touchscreen questionnaires. Participants’ ages were determined based on their reported date of birth. Smoking status (never, former, current), alcohol consumption frequency (never, former, current), and family history of colorectal cancer (yes/no) were all self-reported by participants through the baseline questionnaire. The Townsend Deprivation Index (TDI), which serves as a measure of socioeconomic status, was provided by UKB. Regular physical activity was defined as ≥ 150 min per week of walking or moderate-intensity activity, or ≥ 75 min per week of vigorous-intensity activity, or an equivalent combination of both [12]. A healthy diet was defined as increasing the consumption of fruits, vegetables, whole grains, and fish, while reducing the intake of red meat and processed meats [12]. Physical measurements (such as weight, height, and blood pressure) were conducted by trained staff. Body Mass Index (BMI, kg/m²) was calculated as weight (in kilograms) divided by the square of height (in meters). The personal history of diabetes, hypertension, hyperlipidemia, and cardiovascular disease was determined through self-reported diagnoses or medication use, medical records, or through levels of HbA1c, blood glucose, blood pressure, and blood lipids that met diagnostic criteria (Table S1). The estimated glomerular filtration rate based on cystatin C (eGFRcyst) at recruitment were calculated by the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) equations [13].
Statistical analysis
The baseline characteristics of continuous variables were presented as median [interquartile range (IQR)], while categorical variables were reported as frequency (percentage). Kruskal–Wallis tests and Chi-square test were used to compare the differences between the CRC group and the non-CRC group for continuous and categorical variables, respectively. Serum urate levels were log2-transformed to address the right-skewed distribution. HRs represent the change in CRC risk per 1-unit increase in log₂-transformed urate levels, which corresponds to a doubling of the original urate concentration. We used multivariable Cox proportional hazards regression models to estimate the adjusted hazard ratios (HRs) and 95% confidence intervals (95% CI) for the association between serum urate levels and incident CRC risk. In our analysis, we considered potential covariates that could affect the relationship between urate levels and CRC outcomes, as suggested by previous studies. We reported results from three models: Model 1 adjusted for age and sex; Model 2 further adjusted for study center, BMI, education level, TDI, smoking status, alcohol consumption, family history of CRC, physical activity, and dietary habits; Model 3 additionally adjusted for baseline comorbidities (including diabetes, hypertension, hyperlipidemia, coronary heart disease), eGFRcyst, and history of aspirin use. Given the impact of sex on serum urate concentrations and CRC risk, subgroup analyses were conducted by sex. We also performed subgroup analyses by CRC subsite (colon cancer, rectal cancer). To assess the potential non-linear relationship between serum urate levels and CRC risk, we used restricted cubic splines (RCS) with 3 knots placed at the 10th, 50th, and 90th percentiles of the distribution. All statistical analyses were performed in R software, version 4.2.3.
Two-sample Mendelian randomization
Genetic instrument for urate
As shown in Table S4, the instrumental variables for urate were obtained from the genome-wide association study [14] (accession no. GCST90014015) and exome sequencing study [15] (accession no. GCST90025965) based on UKB cohort. Additionally, we sourced summary data for uric acid levels from the MRC Integrative Epidemiology Unit (IEU) Open GWAS database (accession no. bbj-a-57) [16].
The following criteria were applied to select appropriate SNPs as valid IVs. To ensure compliance with MR assumptions, we selected instrumental SNPs strongly associated with serum urate levels, with a significance threshold of P < 5 × 10− 8. We also applied a conservative clumping threshold of r2 < 0.001 to maintain linkage equilibrium. Furthermore, the strength of each SNP was assessed using the F-statistic; SNPs with an F-statistic value greater than 10 were deemed to have a strong predictive potential for urate levels. Based on the three studies mentioned, 314, 42 and 266 SNPs were identified as IVs for urate or uric acid levels and were used in the MR analysis (Table S5).
GWAS summary-level data of CRC
We selected two CRC-GWAS datasets as outcome data, with details and relevant information provided in Table S4. The first CRC GWAS summary data was derived from a meta-analysis of GWAS [14] from BioBank Japan, UKB, and FinnGen, including 14,886 CRC patients and 622,807 controls (accession no. GCST90018808). Additionally, we obtained a CRC GWAS summary dataset from the IEU Open GWAS database to replicate our findings (accession no. ieu-b-4965).
Two-sample MR analysis
In MR analysis, urate levels were considered the exposure variable, with CRC as the outcome. The inverse variance-weighted (IVW) method, known for providing unbiased estimates, was employed in our MR analysis. Utilizing IVW, we derived the odds ratio (OR) to assess the magnitude of the causal effect, including its 95% confidence interval (CI).
We used the fixed-effect IVW method as our primary approach in the MR analysis to derive an overall weighted estimate of the potential causal effect of serum urate levels on the risk of CRC. In addition, we performed MR-Egger, weighted mode, weighted median, and simple mode analyses to ensure the robustness of our results.
To test the robustness of our results, several sensitivity analyses were conducted, including Cochran’s Q test for heterogeneity. Horizontal pleiotropy was further assessed using MR-Egger regression. Moreover, a leave-one-out analysis was conducted to evaluate whether the causal association was influenced by any single SNP. Statistical significance was defined as P-values being less than 0.05 and all tests were two-sided. MR analyses were conducted using R version 4.2.3 with the ‘TwoSampleMR’ package (version 0.5.6).
Results
Baseline characteristics of prospective cohort study
A total of 180,480 participants were included in this study, of whom, 1960 participants were diagnosed with CRC during the follow-up period. The baseline characteristics of the CRC cases group and control group are detailed in Table 1. Compared to the controls, CRC patients were more likely to be male, older, smoker, had a higher BMI, lower educational attainment, lower levels of eGFRcyst and higher urate levels (all P < 0.05). Furthermore, CRC patients were more likely to have a family history of CRC and to suffer from comorbidities, including cardiovascular diseases, diabetes, hypertension, and hyperlipidemia (all P < 0.05). No significant differences were observed between the CRC cases and controls in terms of TDI, alcohol consumption, physical activity, or dietary habits (all P > 0.05).
Table 1.
Baseline characteristics of participants included from UK biobank
| Variables | Non cases (n = 178520) |
CRC cases (n = 1960) |
P |
|---|---|---|---|
| Age at recruitment | 56 (49, 62) | 61 (55, 65) | < 0.001 |
| Male | 86,701 (48.57) | 1150 (58.67) | < 0.001 |
| BMI, kg/m3 | 26.41 (23.95, 29.41) | 27.20 (24.73, 30.11) | < 0.001 |
| Education levels, College or higher | 72,917 (40.85) | 755 (38.52) | 0.039 |
| TDI | − 2.53 (− 3.85, − 0.35) | − 2.68 (− 3.93, − 0.39) | 0.081 |
| Smoking status | < 0.001 | ||
| Never | 103,943 (58.22) | 973 (49.64) | |
| Former | 61,028 (34.19) | 842 (42.96) | |
| Current | 13,549 (7.59) | 145 (7.40) | |
| Alcohol drinking | 0.219 | ||
| Never | 4431 (2.48) | 47 (2.40) | |
| Former | 4784 (2.68) | 65 (3.32) | |
| Current | 169,305 (94.84) | 1848 (94.29) | |
| Regular physical activity | 148,167 (83.00) | 1605 (81.89) | 0.204 |
| Healthy diet, yes | 40,972 (22.95) | 432 (22.04) | 0.354 |
| Family history of CRC | 19,400 (10.87) | 296 (15.10) | < 0.001 |
| Prevalent diabetes | 8326 (4.66) | 174 (8.88) | < 0.001 |
| Prevalent hypertension | 98,063 (54.93) | 1296 (66.12) | < 0.001 |
| Prevalent hyperlipidemia | 125,459 (70.28) | 1557 (79.44) | < 0.001 |
| Prevalent CVD | 9286 (5.20) | 148 (7.55) | < 0.001 |
| Aspirin use | 22,552 (12.63) | 343 (17.50) | < 0.001 |
| eGFRcyst, mL/min/1.73m2 | 91.77 (80.27, 102.48) | 86.32 (76.41, 97.67) | < 0.001 |
| Urate, µmol/L | 301.30 (248.98, 358.70) | 315.90 (261.15, 370.42) | < 0.001 |
Continuous variables with abnormal distribution were presented as median (IQR). Categorical variables were presented as n (%)
BMI, body mass index; TDI, Townsend deprivation index; CVD, cardiovascular disease
Serum urate levels and incident CRC risk
The associations between urate levels and CRC risk are presented in Table 2. A significant inverse relationship was observed between higher urate levels and CRC risk after adjusting for multiple factors using proportional Cox regression models. Specifically, in the fully adjusted Model 3, each 1-unit increase in log₂-transformed urate levels was associated with a 15% decrease in incident CRC risk (HR: 0.85; 95% CI: 0.72–0.99; P = 0.038). When participants were stratified by cancer type, a significant association between higher urate levels and decreased risk of colon cancer was also observed (Model 3 HR: 0.79; 95% CI: 0.65–0.96; P = 0.015). However, no statistically significant associations were found between serum urate levels and the risk of rectal cancer in the fully adjusted model (Table 2. Additionally, we further stratified the participants by sex, and the results are presented in Table 2. A significant association between higher levels of urate and decreased risk of CRC (Model 3 HR : 0.79; 95% CI: 0.62–1.00; P = 0.050) and colon cancer (HR: 0.66; 95% CI: 0.50–0.87; P = 0.003) was observed in female participants. However, no statistically significant associations were found between serum urate levels and the risk of either colon cancer or rectal cancer in male participants. The heterogeneity analyses revealed a subsite difference in the association of urate with the risk of colon and rectal cancer in women (P-heterogeneity = 0.019), and no other sex and subsite differences were observed (all P-heterogeneity > 0.05). These associations remained when excluding cases that occurred within 1 year of follow-up (Table S3). Using multivariable RCS to explore nonlinear associations, no evidence of deviation from linearity was found. The risk of CRC, colon cancer, and female colon cancer all decreased monotonically with increasing urate levels (all P for overall association < 0.05, but all P for nonlinear association > 0.05, Fig. 1).
Table 2.
Association of serum urate with incident risks of CRC.
| Model | Colorectal cancer | Colon cancer | Rectal cancer | P heterogeneity b | |||
|---|---|---|---|---|---|---|---|
| HR (95%CI) a | P | HR (95%CI) a | P | HR (95%CI) a | P | ||
| All participants | |||||||
| Model 1 | 0.99 (0.86, 1.14) | 0.912 | 0.99 (0.83, 1.18) | 0.942 | 0.98 (0.77, 1.26) | 0.899 | |
| Model 2 | 0.83 (0.72, 0.97) | 0.018 | 0.81 (0.67, 0.97) | 0.026 | 0.87 (0.67, 1.13) | 0.309 | |
| Model 3 | 0.85 (0.72, 0.99) | 0.038 | 0.79 (0.65, 0.96) | 0.015 | 0.97 (0.73, 1.27) | 0.802 | 0.235 |
| Female | |||||||
| Model 1 | 1.00 (0.81, 1.23) | 0.986 | 0.94 (0.74, 1.21) | 0.647 | 1.14 (0.78, 1.67) | 0.497 | |
| Model 2 | 0.85 (0.68, 1.07) | 0.168 | 0.76 (0.58, 0.99) | 0.041 | 1.14 (0.75, 1.74) | 0.530 | |
| Model 3 | 0.79 (0.62, 1.00) | 0.050 | 0.66 (0.50, 0.87) | 0.003 | 1.23 (0.79, 1.92) | 0.357 | 0.019 |
| Male | |||||||
| Model 1 | 1.01 (0.83, 1.24) | 0.908 | 1.07 (0.83, 1.38) | 0.609 | 0.92 (0.66, 1.28) | 0.629 | |
| Model 2 | 0.85 (0.69, 1.04) | 0.121 | 0.89 (0.69, 1.16) | 0.395 | 0.78 (0.55, 1.09) | 0.143 | |
| Model 3 | 0.91 (0.74, 1.14) | 0.422 | 0.94 (0.71, 1.24) | 0.649 | 0.87 (0.61, 1.25) | 0.456 | 0.755 |
| P heterogeneity c | 0.372 | 0.078 | 0.236 | ||||
The colorectal cancer subsite analyses included 1292 new colon cancer cases and 667 new rectal cancer cases, respectively
aCox proportional hazard models were used to obtain the HR (95% CI). Model 1 were adjusted for birth year, gender. Model 2 further adjusted for BMI, education, smoking, drinking, UKB center, TDI, healthy diet, regular physical activity, and family history of CRC. Model 3 further adjusted for diabetes, hypertension, hyperlipidemia, CVD, eGFRcyst and aspirin use. Sex-stratified analyses were performed without adjusting for sex
bP for heterogeneity between colon cancer and rectal cancer was derived from Cochran’s Q test of heterogeneity on the Model 3
cP for heterogeneity between men and women was derived from Cochran’s Q test of heterogeneity on the Model 3
Fig. 1.
Association between urate and colorectal cancer based on the restricted cubic spline regression model. The minimum values of log2-transformed serum levels of urate were used as the reference. Model were adjusted for birth year, gender, UKB center, BMI, education levels, TDI, smoking status, alcohol drinking status, healthy diet, regular physical activity, family history of CRC, eGFRcyst, aspirin use, and history of diseases (diabetes, hypertension, hyperlipidemia, CVD)
Mendelian randomization results
We performed Mendelian randomization (MR) analyses to explore the potential causal link between urate levels and colorectal cancer, with both analyses yielding concordant results in the same direction. In the first MR analysis, using the exposure settings ebi-a-GCST90025965, bbj-a-57, ebi-a-GCST90014015, and the outcome ebi-a-GCST90018808 (Table 3), The IVW method indicated that elevated serum urate levels were significantly associated with a reduced risk of CRC. The ORs and their corresponding 95% CIs were as follows: OR = 0.930 (95% CI: 0.878–0.986, P = 0.014), OR = 0.898 (95% CI: 0.808–0.998, P = 0.047), and OR = 0.857 (95% CI: 0.774–0.948, P = 0.003).
Table 3.
MR analysis result of urate on colorectal cancer risk
| Exposure | Outcome | β | SE | OR | 95%CI | P |
|---|---|---|---|---|---|---|
|
ebi-a-GCST90025965 bbj-a-57 ebi-a-GCST90014015 |
ebi-a-GCST90018808 | − 0.072 | 0.029 | 0.930 | 0.878–0.986 | 0.014 |
| − 0.107 | 0.054 | 0.898 | 0.808–0.998 | 0.047 | ||
| − 0.154 | 0.052 | 0.857 | 0.774–0.948 | 0.003 | ||
| ieu-b-4965 | − 0.001 | 0.0006 | 0.998 | 0.997–0.999 | 0.011 | |
| − 0.001 | 0.0008 | 0.998 | 0.997–1.0004 | 0.160 | ||
| − 0.002 | 0.0010 | 0.997 | 0.995–0.999 | 0.004 |
In the second MR analysis, utilizing exposure settings ebi-a-GCST90025965, bbj-a-57, and ebi-a-GCST90014015, and the outcome ieu-b-4965 as detailed in Table 3, the IVW method demonstrated that elevated serum urate levels were significantly associated with a lower risk of CRC. The ORs and their corresponding 95%CIs were as follows: OR = 0.998 (95% CI: 0.997–0.999, P = 0.011), OR = 0.998 (95% CI: 0.997–1.0004, P = 0.160), and OR = 0.997 (95% CI: 0.995–0.999, P = 0.004). Overall, the MR analysis established a significant causal association between serum urate levels and CRC. Figures 2 and 3 present scatter plots that depict the estimated causal relationship between serum urate levels and CRC, as determined by the IVW method.
Fig. 2.
Visualization of Mendelian randomization (MR) estimates showing the causal effects of genetically predicted colorectal cancer on urate based on the scatter plots
Fig. 3.
Visualization of Mendelian randomization (MR) estimates showing the causal effects of genetically predicted colorectal cancer on urate based on the funnel plots
To assess the robustness of our MR results, we performed a leave-one-out sensitivity analysis. Additionally, we performed horizontal pleiotropy tests to assess whether the SNPs utilized as IVs displayed horizontal pleiotropic effects. In the initial MR analysis, no indication of horizontal pleiotropy was detected (P > 0.05). Similarly, the subsequent MR analysis revealed no signs of horizontal pleiotropy (P > 0.05) These horizontal pleiotropy test results suggest that the selected SNPs, as IVs, influence CRC risk primarily through their effect on urate levels.
Discussion
By integrating a large population-based cohort study with MR analyses, our research delved into the association between urate levels and CRC outcomes. The findings indicate that serum urate levels, when elevated, are associated with a lower risk of CRC incidence. Furthermore, the MR analyses provided evidence supporting the causal nature of these associations.
A UK cohort study involving 444,462 participants indicated that elevated serum uric acid levels are linked to a higher risk of pancreatic and renal cancer in females, as well as gallbladder cancer in males [17, 18]. Additionally, two prospective studies within the Korean population have identified a positive correlation between elevated serum uric acid levels and the risk of developing various cancers, including prostate, esophagus, stomach, liver, pancreatic, lung, ovarian, renal, and bladder cancers [19, 20]. Similarly, cohort studies have reported an increased risk of CRC in Swedish [21] and Chinese males [22]. However, recent studies on the correlation between serum uric acid levels and cancer risk have revealed inconsistent findings within the scientific literature [23]. While some research suggests that elevated serum uric acid levels are linked to a reduced mortality rate across all cancer types [24], other investigations have shown a lower risk of CRC in individuals with higher serum uric acid levels compared to those with lower levels [25]. The discrepancies observed across various studies could stem from factors like population diversity, methodological variations, differing outcome criteria, and the influence of potential confounding factors. In this study, we found an inverse association between urate levels and CRC incidence risk. Although unadjusted comparisons suggested higher urate levels in CRC cases, multivariable adjustment revealed an inverse association, emphasizing the profound confounding by BMI, diabetes, and dietary factors. This paradox may reflect urate’s dual role as both an antioxidant and a marker of metabolic dysfunction, with its net effect depending on the underlying health context [26]. A study based on the Chinese Health and Nutrition Survey found a nonlinear relationship between uric acid and insulin resistance only in overweight/obese nondiabetic individuals [27]. Future studies should prioritize stratified analyses to identify subgroups where urate’s protective or harmful effects predominate. Furthermore, the overall negative association between urate and CRC risk was driven primarily by colon cancer, with no significant correlation observed for rectal cancer. This divergence may reflect biological differences between colon and rectal carcinogenesis or limited statistical power for rectal cancer due to fewer cases.
Notably, uric acid has been demonstrated to have a complex relationship with the inflammatory response [28, 29]. It can directly activate mitogen-activated protein kinases (MAPKs), initiate inflammatory signaling pathways, and increase the production of pro-inflammatory cytokines such as interleukin 1 (IL-1), interleukin 6 (IL-6), and tumor necrosis factor (TNF). These cytokines exert pro-inflammatory effects that are independent of reactive oxygen species (ROS) [30, 31]. Serum uric acid is capable of upregulating mRNA expression of inflammasome-associated genes, including IL-1β and NLRP3, mediated by monosodium urate crystals and soluble uric acid, which may potentially contribute to oncogene activation [32, 33]. However, studies also suggest that uric acid, being the end product of human purine metabolism, serves as a potent antioxidant [34, 35]. It plays a significant role in human physiology, accounting for a substantial portion of the scavenging activity against free radicals [36]. Furthermore, uric acid functions as a primary endogenous danger signal emitted by damaged cells, capable of stimulating dendritic cell maturation and augmenting CD8 + T cell response, thereby reinforcing immune responses [37, 38]. As an antioxidant, uric acid has been reported to have a protective role against cancer [39, 40].
The strengths of our research are notable, particularly the large sample size of CRC cases, which enhances statistical power and allows for the application of diverse methods to assess potential biases related to pleiotropy. Furthermore, our MR study leverages a genetically predicted phenotype as the exposure, thereby diminishing the potential for reverse causality and confounding bias associated with conventional observational research. Using SNPs as instrumental variables for exposure, MR analysis offers a methodology akin to randomized controlled trials, given the random distribution of these genetic variants among individuals [41]. Our MR study’s findings indicate that elevated urate levels might offer clinical benefits by diminishing the risk of CRC outcomes, with substantial public health implications. Finally, we conducted replication analyses on two separate datasets, significantly enhancing the reliability and trustworthiness of our findings. Our study has several limitations that warrant consideration. To minimize potential bias associated with ethnic diversity, we limited our participant pool to individuals of European ancestry. Nonetheless, it is crucial to validate our findings in diverse populations to ascertain their broader applicability. Our analyses excluded close to half of the cohort participants, primarily due to missing covariate data. Thus, there is a possibility of selection bias that potentially affects the generalization of our results. Future studies should prioritize different cohorts and advanced missing data methods. Furthermore, UKB participants tend to be healthier than the general population, and smoking-related disease rates are significantly lower, which may introduce selection bias [42]. Lastly, our MR analysis was limited by the available SNPs. Expanding our scoring model to include a broader range of urate-related SNPs could strengthen our causal inference. We acknowledge potential correlations between some SNPs and unidentified CRC factors, necessitating further validation as more GWAS data emerge.
Conclusion
This study is the first large-scale effort to explore the link between urate levels and CRC, using both cohort and MR analyses, which bolsters the reliability of our findings.
Supplementary Information
Below is the link to the electronic supplementary material.
Author contributions
YH.Z. and YJ.L. and N.S. wrote the main manuscript text and KM.X. and H.H. and QW.B.prepared Figs. 1, 2 and 3. All authors reviewed the manuscript.
Funding
Not applicable.
Data availability
Data is provided within the manuscript or supplementary information files.
Declarations
Ethics approval and consent to participate
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yuhan Zhou and Kemin Xu contributed equally to this work.
Contributor Information
Na Shen, Email: shenna@tjh.tjmu.edu.cn.
Yanjun Lu, Email: yanjunlu@tjh.tjmu.edu.cn.
References
- 1.Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin. 2023;73(1):17–48. [DOI] [PubMed] [Google Scholar]
- 2.Fang Z, Giovannucci EL. The timing of adiposity and changes in the life course on the risk of cancer. Cancer Metastasis Rev. 2022;41(3):471–89. [DOI] [PubMed] [Google Scholar]
- 3.Mayen AL, Viallon V, Botteri E, Proust-Lima C, Bagnardi V, Batista V, Cross AJ, Laouali N, MacDonald CJ, Severi G, et al. A longitudinal evaluation of alcohol intake throughout adulthood and colorectal cancer risk. Eur J Epidemiol. 2022;37(9):915–29. [DOI] [PubMed] [Google Scholar]
- 4.Kennedy J, Alexander P, Taillie LS, Jaacks LM. Estimated effects of reductions in processed meat consumption and unprocessed red meat consumption on occurrences of type 2 diabetes, cardiovascular disease, colorectal cancer, and mortality in the USA: a microsimulation study. Lancet Planet Health. 2024;8(7):e441–51. [DOI] [PubMed] [Google Scholar]
- 5.Giovannucci E, Michaud D. The role of obesity and related metabolic disturbances in cancers of the colon, prostate, and pancreas. Gastroenterology. 2007;132(6):2208–25. [DOI] [PubMed] [Google Scholar]
- 6.Collaborators GBDRF. Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2020;396(10258):1223–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Strasak AM, Rapp K, Hilbe W, Oberaigner W, Ruttmann E, Concin H, Diem G, Pfeiffer KP, Ulmer H. Vhm et al: serum uric acid and risk of cancer mortality in a large prospective male cohort. Cancer Causes Control. 2007;18(9):1021–9. [DOI] [PubMed] [Google Scholar]
- 8.Liao W, Wang Y, Zhang W. Serum uric acid and the risk of colorectal cancer: a meta-analysis. Eur J Cancer Prev. 2024;33(1):19–28. [DOI] [PubMed] [Google Scholar]
- 9.Horsfall LJ, Nazareth I, Petersen I. Serum uric acid and the risk of respiratory disease: a population-based cohort study. Thorax. 2014;69(11):1021–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Pan C, Cheng B, Qin X, Cheng S, Liu L, Yang X, Meng P, Zhang N, He D, Cai Q et al. Enhanced polygenic risk score incorporating gene-environment interaction suggests the association of major depressive disorder with cardiac and lung function. Brief Bioinform. 2024;25(2):1-9. [DOI] [PMC free article] [PubMed]
- 11.Emdin CA, Khera AV, Natarajan P, Klarin D, Zekavat SM, Hsiao AJ, Kathiresan S. Genetic association of Waist-to-Hip ratio with cardiometabolic Traits, type 2 Diabetes, and coronary heart disease. JAMA. 2017;317(6):626–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zhu M, Wang T, Huang Y, Zhao X, Ding Y, Zhu M, Ji M, Wang C, Dai J, Yin R, et al. Genetic risk for overall cancer and the benefit of adherence to a healthy lifestyle. Cancer Res. 2021;81(17):4618–27. [DOI] [PubMed] [Google Scholar]
- 13.Inker LA, Schmid CH, Tighiouart H, Eckfeldt JH, Feldman HI, Greene T, Kusek JW, Manzi J, Van Lente F, Zhang YL, et al. Estimating glomerular filtration rate from serum creatinine and Cystatin C. N Engl J Med. 2012;367(1):20–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sakaue S, Kanai M, Tanigawa Y, Karjalainen J, Kurki M, Koshiba S, Narita A, Konuma T, Yamamoto K, Akiyama M, et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet. 2021;53(10):1415–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Barton AR, Sherman MA, Mukamel RE, Loh PR. Whole-exome imputation within UK biobank powers rare coding variant association and fine-mapping analyses. Nat Genet. 2021;53(8):1260–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kanai M, Akiyama M, Takahashi A, Matoba N, Momozawa Y, Ikeda M, Iwata N, Ikegawa S, Hirata M, Matsuda K, et al. Genetic analysis of quantitative traits in the Japanese population links cell types to complex human diseases. Nat Genet. 2018;50(3):390–400. [DOI] [PubMed] [Google Scholar]
- 17.Huang CF, Huang JJ, Mi NN, Lin YY, He QS, Lu YW, Yue P, Bai B, Zhang JD, Zhang C, et al. Associations between serum uric acid and hepatobiliary-pancreatic cancer: A cohort study. World J Gastroenterol. 2020;26(44):7061–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Dai XY, He QS, Jing Z, Yuan JQ. Serum uric acid levels and risk of kidney cancer incidence and mortality: A prospective cohort study. Cancer Med. 2020;9(15):5655–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kim YR, Choi CK, Lee YH, Choi SW, Kim HY, Shin MH, Kweon SS. Association between Albumin, total Bilirubin, and uric acid serum levels and the risk of cancer: A prospective study in a Korean population. Yonsei Med J. 2021;62(9):792–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Oh YJ, Lee YJ, Lee E, Park B, Kwon JW, Heo J, Moon KW. Cancer risk in Korean patients with gout. Korean J Intern Med. 2022;37(2):460–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Yiu A, Van Hemelrijck M, Garmo H, Holmberg L, Malmstrom H, Lambe M, Hammar N, Walldius G, Jungner I, Wulaningsih W. Circulating uric acid levels and subsequent development of cancer in 493,281 individuals: findings from the AMORIS study. Oncotarget. 2017;8(26):42332–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Li W, Liu T, Siyin ST, Zhang Q, Wang Y, Cao L, Qu J. The relationship between serum uric acid and colorectal cancer: a prospective cohort study. Sci Rep. 2022;12(1):16677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhao X, Ding N. Investigating the causal association between serum uric acid levels and gastric cancer risk: a Mendelian randomization study. Sci Rep. 2024;14(1):26165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Taghizadeh N, Vonk JM, Boezen HM. Serum uric acid levels and cancer mortality risk among males in a large general population-based cohort study. Cancer Causes Control. 2014;25(8):1075–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hu KY, Cheng YQ, Shi ZL, Ren FP, Xiao GF. Casual associations between blood metabolites and colon cancer. World J Gastrointest Oncol. 2024;16(5):1995–2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wen SJ, Arakawa H, Tamai I. Uric acid in health and disease: from physiological functions to pathogenic mechanisms. Pharmacol Ther. 2024;256:108615. [DOI] [PubMed] [Google Scholar]
- 27.Zhang R, Garstka MA, Zhang CH, Ding SM, Xu J. Impact of body mass index on the association between kidney function and insulin resistance in diabetic and nondiabetic populations. J Diabetes Investig. 2025;16(7):1292-1304. [DOI] [PMC free article] [PubMed]
- 28.Tan X, Zhang Y, Luo P, Lin Z, Li F, Liu H. Fucoidan from laminaria Japonica protects renal tubular epithelial cells from uric acid induced NLRP3-mediated pyroptosis through Inhibition of NF-kappaB pathway. J Ethnopharmacol. 2024;335:118614. [DOI] [PubMed] [Google Scholar]
- 29.Ruggiero C, Cherubini A, Ble A, Bos AJ, Maggio M, Dixit VD, Lauretani F, Bandinelli S, Senin U, Ferrucci L. Uric acid and inflammatory markers. Eur Heart J. 2006;27(10):1174–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Huang W, Zhang M, Qiu Q, Zhang J, Hua C, Chen G, Xie H. Metabolomics of human umbilical vein endothelial cell-based analysis of the relationship between hyperuricemia and dyslipidemia. Nutr Metab Cardiovasc Dis. 2024;34(6):1528–37. [DOI] [PubMed] [Google Scholar]
- 31.Yang S, Liu H, Fang XM, Yan F, Zhang Y. Signaling pathways in uric acid homeostasis and gout: from pathogenesis to therapeutic interventions. Int Immunopharmacol. 2024;132:111932. [DOI] [PubMed] [Google Scholar]
- 32.Cabau G, Crisan TO, Kluck V, Popp RA, Joosten LAB. Urate-induced immune programming: consequences for gouty arthritis and hyperuricemia. Immunol Rev. 2020;294(1):92–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Oguz N, Kirca M, Cetin A, Yesilkaya A. Effect of uric acid on inflammatory COX-2 and ROS pathways in vascular smooth muscle cells. J Recept Signal Transduct Res. 2017;37(5):500–5. [DOI] [PubMed] [Google Scholar]
- 34.Wan Y, Kang Q, Wang Y, Lu X, Li J, Nie H, Chai W. Association of serum uric acid levels with risk of epilepsy: A National population-based study. Epilepsy Behav. 2024;159:109991. [DOI] [PubMed] [Google Scholar]
- 35.Richette P, Bardin T. Gout Lancet. 2010;375(9711):318–28. [DOI] [PubMed] [Google Scholar]
- 36.Zhang P, Wang R, Qu Y, Guo ZN, Zhen Q, Yang Y. Serum uric acid levels and outcome of acute ischemic stroke: a Dose-Response Meta-analysis. Mol Neurobiol. 2024;61(3):1704–13. [DOI] [PubMed] [Google Scholar]
- 37.Shi Y, Evans JE, Rock KL. Molecular identification of a danger signal that alerts the immune system to dying cells. Nature. 2003;425(6957):516–21. [DOI] [PubMed] [Google Scholar]
- 38.Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflammation. Nature. 2008;454(7203):436–44. [DOI] [PubMed] [Google Scholar]
- 39.Xue X, Sun Z, Ji X, Lin H, Jing H, Yu Q. Associations between serum uric acid and breast cancer incidence: A systematic review and meta-analysis. Am J Med Sci. 2024;368(6):610–20. [DOI] [PubMed] [Google Scholar]
- 40.Ihira H, Nakano S, Yamaji T, Katagiri R, Sawada N, Inoue M, Tsugane S, Iwasaki M. Plasma albumin, bilirubin, and uric acid and the subsequent risk of cancer: a case-cohort study in the Japan public health Center-based prospective study. Am J Epidemiol. 2024;193(10):1460–9. [DOI] [PubMed] [Google Scholar]
- 41.Zhou H, Shen J, Fang W, Liu J, Zhang Y, Huang Y, Zhang L. Mendelian randomization study showed no causality between Metformin use and lung cancer risk. Int J Epidemiol. 2020;49(4):1406–7. [DOI] [PubMed] [Google Scholar]
- 42.Fry A, Littlejohns TJ, Sudlow C, Doherty N, Adamska L, Sprosen T, Collins R, Allen NE. Comparison of sociodemographic and Health-Related characteristics of UK biobank participants with those of the general population. Am J Epidemiol. 2017;186(9):1026–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data is provided within the manuscript or supplementary information files.