Higher serum uric acid linked to lower gastric cancer risk, study suggests

Genetically elevated serum uric acid is associated with a reduced risk of gastric cancer, according to a multidimensional study published in Frontiers in Medicine that combines Mendelian randomization, tumor database analysis and laboratory experiments on gastric cancer cells.

The causal question has been difficult to settle with observational studies alone. Previous meta-analyses have reported that higher serum uric acid is positively associated with liver and colorectal cancer but negatively associated with lung cancer, and a systematic review found elevated uric acid levels in gastric cancer patients despite substantial heterogeneity between studies. One 2024 Mendelian randomization study reported no causal relationship between serum uric acid and gastric cancer, while another study described a nonlinear, U-shaped association in which gastric cancer incidence first fell and then rose as uric acid levels increased. Because observational findings are vulnerable to confounding by renal function, metabolic syndrome, diet, and changes in uric acid caused by the cancer itself or its treatment, as well as reverse causality, the research team turned to genetic instruments that are fixed at conception.

Mendelian randomization uses single nucleotide polymorphisms that are strongly associated with a modifiable exposure, in this case serum uric acid, as instrumental variables to test whether the exposure causally influences an outcome. The researchers drew gastric cancer genome-wide association data from a European population of 476,116 individuals and serum uric acid data from an East Asian population of 129,405 individuals, both derived from a cross-population GWAS analysis of 220 human phenotypes. After screening variants at a significance threshold of p < 5 × 10⁻⁸ and pruning for linkage disequilibrium using a European reference panel, the analysis retained 59 instrumental SNPs with strong instrument strength, with per-SNP F-statistics ranging from 30.00 to 1,409.80 and an overall F-statistic of 174.82, well above the conventional threshold that indicates absence of weak instruments.

The primary inverse-variance weighted analysis produced a statistically significant protective estimate: an odds ratio of 0.76 with a 95 percent confidence interval of 0.62 to 0.92 and a p-value of 0.006, meaning that genetically predicted higher serum uric acid was associated with roughly a 24 percent lower odds of gastric cancer. Four complementary estimators, including MR-Egger, weighted median, simple mode and weighted mode, all pointed in the same negative direction, with odds ratios between 0.76 and 0.98, but only the inverse-variance weighted result reached conventional statistical significance. The authors acknowledge this asymmetry in the strength of evidence across methods as a limitation of the analysis.

Sensitivity testing revealed both reassurance and caution. Cochran’s Q-test indicated significant heterogeneity among the SNP estimates (Q = 109.76, p = 3.5 × 10⁻⁵), and the funnel plot showed an asymmetric shape, yet the MR-Egger intercept test found no evidence of directional horizontal pleiotropy (intercept = −0.0085, p = 0.185), meaning the genetic variants did not appear to influence gastric cancer through pathways independent of uric acid. The MR-PRESSO procedure identified two pleiotropic outlier SNPs, rs80292863 and rs9468498, and the distortion test indicated that removing these outliers did not significantly alter the causal estimate. The outlier-corrected estimate remained significant, with an odds ratio of 0.88 (95 percent CI: 0.77–1.00, p = 0.048), and leave-one-out analysis produced estimates that clustered tightly together, indicating that no single SNP was driving the result.

To probe the biology behind the statistical association, the team mapped the instrumental SNPs to genes and identified 36 candidate genes that may connect uric acid to gastric cancer, including ABCG2, GCKR, GNAS, HNF1A, IGF1R, SLC2A9 and PNPLA3. Using RNA-seq data from The Cancer Genome Atlas, they found that 25 of these genes were differentially expressed between gastric cancer tissue and adjacent normal tissue, with 22 upregulated and three downregulated in tumors. Kaplan–Meier survival analysis showed that the expression levels of six genes, ABCG2, BICC1, CFAP73, GNAS, MXD3 and PNPLA3, were significantly associated with overall survival in gastric cancer patients. High expression of BICC1, ABCG2, GNAS and PNPLA3 marked poorer prognosis, whereas high expression of CFAP73 and MXD3 was associated with better survival. Enrichment analysis linked the gene set to uric acid metabolism, carboxylic acid and organic anion transport, insulin-like growth factor binding activity, and the mTOR signaling pathway, with cellular localization concentrated in the brush border membrane and apical plasma membrane, structures relevant to the epithelial polarity changes seen in gastric carcinogenesis.

The laboratory arm of the study tested whether uric acid directly affects gastric cancer cells. AGS cells, a single gastric cancer cell line, were treated with uric acid at concentrations of 0, 50, 100 and 200 micromolar for 48 hours. Cell viability, measured with a CCK-8 assay, fell in a dose-dependent manner from approximately 100 percent in the control to about 90 percent at 50 micromolar, 75 percent at 100 micromolar and 65 percent at 200 micromolar, with the two highest concentrations differing significantly from the control. Flow cytometry showed a dose-dependent increase in the late apoptotic and necrotic cell population, rising from 1.83 percent to 2.29 percent and peaking at 100 micromolar, alongside a modest decline in the live cell fraction, which the authors describe as a mild but detectable induction of cell death.

The mechanism the experiments suggest is pro-oxidant rather than antioxidant. Intracellular reactive oxygen species rose dose-dependently, with the proportion of high-ROS cells increasing from 32.13 percent to 58.09 percent after uric acid treatment. At the same time, the activity of superoxide dismutase fell significantly at all tested concentrations, catalase activity declined significantly at 100 and 200 micromolar, and glutathione content became significantly reduced at 200 micromolar, indicating that uric acid depleted the cell’s antioxidant defenses rather than reinforcing them. Western blotting added a signaling dimension: the ratio of phosphorylated AKT to total AKT, a measure of PI3K/AKT pathway activity, dropped progressively with increasing uric acid concentration, from 0.85 at the low dose to 0.53 at the medium dose and 0.36 at the high dose relative to the vehicle control. Because the PI3K/AKT pathway drives cell survival, proliferation and metabolism and is frequently hyperactivated in gastric cancer, its suppression offers a plausible route by which uric acid could restrain tumor cell growth.

The authors propose a preliminary model in which long-term, genetically determined elevation of baseline serum uric acid may protect against the early stages of gastric carcinogenesis, potentially through antioxidant properties or through regulation of genes involved in urate metabolism, transport and epithelial polarity maintenance, while in established tumors uric acid may act in a pro-oxidative direction that induces apoptosis in transformed cells. They note an apparent tension between their in vitro finding of PI3K/AKT inhibition and the bioinformatic enrichment of mTOR signaling pathway activation, suggesting that long-term genetic influences of uric acid-related genes and the acute effects of direct uric acid treatment may involve distinct mechanistic emphases that warrant further exploration.

Several caveats temper the conclusions. The exposure and outcome GWAS datasets came from different ancestral populations, an East Asian exposure dataset and a European outcome dataset, which may partially violate the assumptions of two-sample Mendelian randomization, and the European linkage disequilibrium reference panel used for pruning may not fully reflect East Asian genetic structure. Only the inverse-variance weighted method reached statistical significance, and the analysis cannot capture dynamic changes or nonlinear relationships such as U-shaped associations. The cell experiments were conducted in a single gastric cancer line, and the prognostic findings are retrospective and supportive rather than proof of a protective effect. The authors call for ancestry-matched GWAS datasets, validation in additional cell lines and animal models, and functional gain- and loss-of-function experiments to clarify how the candidate genes mediate the apparent relationship between serum uric acid and gastric cancer risk.

Subject of Research: The causal relationship between serum uric acid levels and gastric cancer risk and its underlying molecular mechanisms

Article Title: Multidimensional analysis explores the correlation and potential molecular mechanisms between serum uric acid and risk of gastric cancer

Article References: Wei, W., Zhao, F., Fu, Q., Shen, X. L., Liao, H., & Di, J. (2026). Multidimensional analysis explores the correlation and potential molecular mechanisms between serum uric acid and risk of gastric cancer. Frontiers in Medicine, 13, Article 1929597. https://doi.org/10.3389/fmed.2026.1929597

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1929597

Keywords: serum uric acid, gastric cancer, Mendelian randomization, oxidative stress, PI3K/AKT pathway, reactive oxygen species, AGS cells, mTOR signaling, genetic instruments, TCGA, apoptosis, urate metabolism