Dietary patterns and gastric cancer risk: A systematic review and meta-analysis

P. Bertuccio, V. Rosato, A. Andreano, M. Ferraroni, A. Decarli, V. Edefonti, C. La Vecchia

Research output: Contribution to journalArticlepeer-review

Abstract

Background: Declines in gastric cancer (GC) incidence and mortality have been related to improvements in diet. It is therefore important to consider dietary patterns. Design: We conducted a systematic review and meta-analysis of the literature through Medline and Embase databases. Results: We identified 16 papers, of these 9 derived dietary patterns through an a posteriori method, 5 through a priori scores, and 2 used both approaches. Eight studies that used the a posteriori approach were considered for the meta-analysis. A favorable role on GC emerged for the 'Prudent/healthy', with an odds ratio (OR) of 0.75 [95% confidence interval (CI): 0.63-0.90], for the highest versus the lowest category. Similar results emerged for separate anatomical subtypes. An unfavorable role on GC emerged for the 'Western/unhealthy' dietary pattern, with an OR of 1.51 (95% CI: 1.21-1.89). This association was weaker for the distal/NOS (not otherwise specified) category (OR = 1.36) compared with the cardia GC (OR = 2.05). Among the a priori scores, the ORs ranged from 0.2 to 0.7 for the favorable and from 1.8 to 6.9 for the unfavorable ones. Conclusion: There is a ~2-fold difference in GC risk between a 'Prudent/healthy' diet-rich in fruits and vegetables, and a 'Western/unhealthy' diet-rich in starchy foods, meat and fats.

Original languageEnglish
Pages (from-to)1450-1458
Number of pages9
JournalAnnals of Oncology
Volume24
Issue number6
DOIs
Publication statusPublished - Jun 2013

Keywords

  • Dietary patterns
  • Gastric cancer
  • Meta-analysis
  • Review
  • Score

ASJC Scopus subject areas

  • Oncology
  • Hematology

Fingerprint

Dive into the research topics of 'Dietary patterns and gastric cancer risk: A systematic review and meta-analysis'. Together they form a unique fingerprint.

Cite this