Jean Ching-Yuan Fan

The age-standardized incidence rate of gastric cancer in the Matsu islands was approximately three-fold compared with that in other Taiwan counties. The high prevalence of H. pylori infection plays a strong predictor for the high incidence of gastric cancer in Matsu. Dietary factors plays a sufficient cause of gastric cancer. The effect of dietary intakes on different stages of gastric neoplasia following the Correa model has been proven based on a Mastu cohort consisted of 2,201 residents participated in a two-stage screening project for gastric neoplasia. The special dietary behavior, such as high intake of salty food, seafood, and frequent intake of liquor were also of high interesting to be investigated their effect on each gastric cancer. Therefore, we used the dietary factor pattern identified in the previous article and combined with alcohol drinking and tea consumption as the predictors for developing each stage of the Correa model. Effects of dietary variables aggregated by factor analysis. were evaluated. Those association were analyzed with the Cox proportional hazards regression model. Only those variables show statistically significant at the P-value less 0.10 will be put into multivariable Cox model. There are 8 factors were selected in factor analysis. In multivariable Cox analysis, Factor 1 (salty food) has borderline harmful effect (HR=1.41;95%CI: 0.95-2.08) on atrophy gastritis (AG), Factor 2 (seafood) shows protecting effect (HR=0.81;95%CI:0.68-0.96) on superficial gastritis (SG). Factor 3 (vegetable &bean) has harmful effect (HR=1.79;95%CI: 1.16-2.77) on AG. Factor 4 (meat) has harmful effect (HR=1.25;95%CI:1.05-1.49) on SG. Alcohol drinking was associated with SG in univariable analysis showing a 2.4 fold for those quitter compared with those without habit of alcohol drinking (HR=2.40; 95% CI: 1.16-4.96), but it was not statistically significant after adjusting other covariates. Tea consumption was not associated with any stage of gastric neoplasia.

In conclusion, dietary habits not only play the role in the development of early precursor (SG and AG) but also make contribution to late precursor (intestinal metaplasia;IM) and gastric cancer.

Keywords: Dietary factor, gastric cancer, Correa model, precursor