Increased risk of early-onset gastric cancer among young adults age 20–49 years with diabetes: a nationwide cohort study

Summary: Background: Although overall gastric cancer (GC) incidence has declined, the incidence of early-onset GC (EOGC) is rising. Given the increasing prevalence of diabetes in younger populations and its link to malignancies, we investigated the association between glycemic status and EOGC risk. Methods: This nationwide cohort study followed 2,439,112 adults (aged 20–49 years) without a prior cancer diagnosis who underwent national health screening in 2012, through 2022. EOGC risk was assessed using multivariable Cox models, with restricted cubic spline and sensitivity analyses (inverse probability weighting [IPW] and exclusion of early diabetes). Findings: The incidence rate (IR) per 1000 person-years rose with worsening glycemic status: 0.23 (95% confidence interval [CI], 0.22–0.23) in individuals with normoglycemia, 0.30 (95% CI, 0.28–0.32) in individuals with impaired fasting glucose, 0.41 (95% CI, 0.34–0.48) in newly diagnosed individuals with diabetes, 0.51 (95% CI, 0.40–0.64) in individuals with diabetes <5 years, and 0.62 (95% CI, 0.48–0.80) in individuals with diabetes ≥5 years. In adjusted models, EOGC risk was higher in newly diagnosed diabetes (adjusted hazard ratio [aHR], 1.27; 95% CI, 1.07–1.52), diabetes <5 years (aHR, 1.36; 95% CI, 1.07–1.74), and diabetes ≥5 years (aHR, 1.53; 95% CI, 1.17–2.01). Results were consistent in spline, IPW, and exclusion analyses. Current smoking and both light-to-moderate and heavy alcohol consumption were associated with increased risk across all glycemic groups, with the highest risk in diabetes. Interpretation: Worsening glycemic status is associated with increased risk of EOGC in a dose-dependent manner. Optimal glycemic control, smoking cessation and reduced alcohol consumption may mitigate EOGC risk. However, residual confounding, missing data on Helicobacter pylori infection, family history, dietary factors, and histological subtypes, together with potential detection bias and limited information on glucose-lowering therapy or longitudinal glycemic status, should be acknowledged as limitations. Funding: None.