Risk factors for gastric cancer: A comprehensive analysis of observational studies

Abstract

Background: Multifarious factors have a causal relationship with gastric cancer (GC) development. We conducted a comprehensive analysis to evaluate the strength of the evidence examining non-genetic risk factors for gastric cancer.

Methods: PubMed, Web of Science, and the Cochrane Library were searched from inception to November 10, 2021 to identify meta-analyses of observational studies examining the association between environmental factors and GC risk. For each meta-analysis, the random effect size, 95% confidence interval, heterogeneity among studies, and evidence of publication bias were assessed; moreover, the evidence was graded using predefined criteria, and the methodological quality was evaluated using AMSTAR 2.

Results: A total of 137 associations were examined in 76 articles. Among these meta-analyses, 93 associations yielded significant estimates (p < 0.05). Only 10 associations had strong epidemiologic evidence, including 2 risk factors (waist circumference and bacon), and 8 protective factors (dietary total antioxidant capacity, vegetable fat, cruciferous vegetable, cabbage, total vitamin, vitamin A, vitamin C, and years of fertility); 26 associations had moderate quality of evidence; and the remaining 57 associations were rated as weak. Ninety-four (68.61%) associations showed significant heterogeneity. Twenty-five (18.25%) associations demonstrated publication bias.

Conclusions: In this comprehensive analysis, multiple associations were found between environmental factors and GC with varying levels of evidence. Healthy dietary habits and lifestyle patterns could reduce the risk for GC. However, further high-quality prospective studies are still necessary to draw more definitive conclusions.

Keywords: comprehensive analysis; gastric cancer; protective factors; quality; risk factors.

Figure 1

Flow diagram for literature search and selection process.

Figure 2

Forest plot: summary effect estimates of meta-analyses reporting associations between GC and factors pertaining to anthropometric indices. BMI, body mass index; HvL, highest vs. lowest; NA, not applicable; red dots represent risk factors; blue dots represent protective factors; The strength of the epidemiologic evidence was rated as high (⊕⊕⊕), moderate (⊕⊕), weak (⊕).

Figure 3

Forest plot: summary effect estimates of meta-analyses reporting associations between GC and factors pertaining to dietary intake. DII, dietary inflammatory index; D-TAC, dietary total antioxidant capacity.

Figure 4

Forest plot: summary effect estimates of meta-analyses reporting associations between GC and factors pertaining to micronutrients.

Figure 5

Forest plot: summary effect estimates of meta-analyses reporting associations between GC and factors pertaining to use of medication. PPI, proton pump inhibitors.

Figure 6

Forest plot: summary effect estimates of meta-analyses reporting associations between GC and factors pertaining to use of lifestyle.

Figure 7

Forest plot: summary effect estimates of meta-analyses reporting associations between GC and factors pertaining to pre-existing medical history. NAFLD, non-alcoholic fatty liver disease; SLE, systemic lupus erythematosus; IBD, inflammatory bowel disease; GDM, gestational diabetes mellitus.

Figure 8

Forest plot: summary effect estimates of meta-analyses reporting associations between GC and factors pertaining to viral or bacterial infection. HP, Helicobacter pylori; EBV, Epstein-Barr virus; HBV, hepatitis B virus; HCV, hepatitis C virus; HCMV, human cytomegalovirus; HPV, human papillomavirus; HTLV-1, human T-cell lymphotropic virus type 1; JCV, John Cunningham virus.

Figure 9

Forest plot: summary effect estimates of meta-analyses reporting associations between GC and factors pertaining to other factors. PM2.5, particulate matter with a diameter of 2.5 μm or less.

Figure 10

Map of results of AMSTAR 2 scores.

Figure 11

Map of results of epidemiologic evidence assessment.