Dietary patterns and risk of gastric cancer: A case-control study in Uruguay

Eduardo De Stefani, Pelayo Correa, Paolo Boffetta, Hugo Deneo-Pellegrini, Alvaro L. Ronco, María Mendilaharsu

Research output: Contribution to journalArticlepeer-review

109 Scopus citations


Background. Gastric cancer is a frequent malignancy in the Uruguayan population. In northern counties, incidence rates reach high figures (age-standardized rates [ASR], 37.3 per 100000 men and 18.3 per 100000 women). Diet is a major determinant in gastric carcniogenesis. Because foods or food groups have the advantage over nutrients in being most directly related to dietary recommendations, we decided to conduct a case-control study on the relationships between food groups and risk of gastric cancer. For this purpose, we included 240 cases and 960 controls. Methods. In the present study we employed three analytical approaches: (1) individual food group analysis, (2) factor analysis, and (3) analysis of empirical scores of risk. Individual analysis of food groups was performed by multiple unconditional logistic regression, with food groups being the explanatory variables. Nineteen food groups were created and categorized in tertiles according to the control distribution. Factor analysis aggregated iutercorrelated foods in broader eating patterns. In this study we were able to identify three factors or patterns, arbitrarily labeled "starchy," "healthy," and "mixed." Finally, empirical scores of risk were created after examining the risk of each individual food and summing each significant item. This resulted in a risk enhancing score and a protective score. Both scores were categorized in tertiles according to the control distribution. Results. The individual analysis of food groups showed increased risks of gastric cancer for rice, salted meat, stewed meat, white bread, potatoes, and tubers. Ou the other hand, raw vegetables, total fruits, legumes, and black tea were inversely associated with risk of gastric cancer. All three dietary patterns, generated by factor analysis, were significantly associated with gastric carcinoma risk. Whereas the starchy factor was directly associated with gastric cancer, the healthy and mixed patterns were strongly protective. Finally, the risk enhancing empirical score displayed an increased risk of gastric cancer (odds ratio [OR], 4.1, 95% confidence interval [CI], 2.6-6.6), whereas the protective score showed aN important reduction in risk, of 0.38. Conclusion. This study displayed consistent results from three different approaches. Concerning different food groups, stewed and processed meat are rich in salt; rice, tubers, and winter squash are sources of starch; and vegetables and fruits are rich in ascorbic acid and carotenoids. All these substances have been strongly related to gastric carcinogenesis. Furthermore, this study suggests that diets rich in vegetables and fruits and with low amounts of salty and starchy foods are recommendable for the prevention of gastric cancer.

Original languageEnglish
Pages (from-to)211-220
Number of pages10
JournalGastric Cancer
Issue number4
StatePublished - Dec 2004
Externally publishedYes


  • Empirical scores
  • Factor analysis
  • Foods
  • Gastric cancer


Dive into the research topics of 'Dietary patterns and risk of gastric cancer: A case-control study in Uruguay'. Together they form a unique fingerprint.

Cite this