Background: Although advances in diagnosis and treatment have improved survival outcomes for gastric cancer, the extent to which regional socioeconomic development influences long-term prognosis across disease stages remains insufficiently understood.
Aim: We examined the association between regional socioeconomic status (SES) and stage-specific survival among patients with gastric cancer in China.
Methods: We analyzed data from the National Gastric Cancer Cohort, comprising 38,557 patients diagnosed between 1998 and 2022 at nine tertiary hospitals nationwide. Median overall survival (OS) was 51.4 months (95% CI, 50.0–52.9). Regional SES was defined using gross domestic product (GDP) per capita and the Human Development Index (HDI), and patients were categorized into high, medium, or low SES groups according to region of residence. Survival was estimated using the Kaplan–Meier method, and multivariable Cox proportional hazards models were used to evaluate associations between SES and all-cause mortality.
Results: Compared with patients from medium–low GDP regions, those from high–GDP regions had higher prevalences of smoking, alcohol consumption, overweight or obesity, and stage IV disease at diagnosis. Among patients with stage I–III disease, those from high–GDP regions were slightly less likely to receive gastrectomy or neoadjuvant therapy but more likely to receive adjuvant therapy; similar patterns were observed when SES was defined by HDI. Five-year OS declined progressively with decreasing SES: 51.1%, 46.2%, and 46.1% across high-, medium-, and low-GDP regions, respectively (P < 0.001), with consistent gradients across stages I-III but not stage IV (stage I: 91.8%, 89.5%, 89.7%; stage II: 73.4%, 66.0%, 68.9%; stage III: 40.8%, 38.8%, 32.1%; stage IV: 16.2%, 10.7%, 13.8%). Using HDI, corresponding 5-year OS rates were 51.3%, 48.5%, and 44.1% (P < 0.001). After multivariable adjustment, residence in medium- or low-SES regions was independently associated with higher mortality compared with high-SES regions, whether defined by GDP per capita (medium: hazard ratio [HR] 1.216, 95% CI 1.156-1.278; low: HR 1.051, 95% CI 1.007-1.096) or HDI (medium: HR 1.154, 95% CI 1.103-1.208; low: HR 1.064, 95% CI 1.018-1.112). Formal interaction tests between SES and TNM stage were not statistically significant; however, stage-stratified analyses indicated that the survival advantage associated with higher SES was largely confined to stage I-III disease.
Conclusion: The stage-specific survival gradients observed in China highlight a broader challenge facing many middle-income countries (MICs) undergoing rapid epidemiologic and socioeconomic transition. As cancer care capacity expands, unequal health-system performance, rather than lack of treatment availability alone, may become a dominant driver of survival disparities, particularly for cancers that are potentially curable with timely, standardized care. These findings underscore the need for MICs to move beyond access-focused policies toward equity-oriented cancer control strategies that integrate early detection, quality-assured treatment pathways, and continuous monitoring of outcomes across regions and population groups. Strengthening data-driven accountability within health systems and prioritizing reduction of avoidable survival gaps should be central to future national cancer control plans and global cancer equity agendas.