Table 2

Summary statistics for the association between H. pylori and esophageal adenocarcinoma

No. studiesQ statistic*PI2 (%)Random effects OR (95% CI)Fixed effects OR (95% CI)
H. pylori (840/2,890)§
 All studies1314.190.29150.56 (0.46–0.68)0.55 (0.47–0.66)
 Large studies (SE < 0.5)89.640.21270.58 (0.47–0.73)0.57 (0.48–0.68)
 Population-based studies67.320.20320.58 (0.43–0.76)0.57 (0.46–0.71)
 Western studies1213.280.28170.57 (0.47–0.70)0.56 (0.47–0.67)
 Eastern studies100.32 (0.10–1.02)0.32 (0.10–1.02)
 Studies used serologic tests1010.030.35100.59 (0.48–0.73)0.56 (0.47–0.67)
 Studies with age-matched controls89.170.24240.58 (0.46–0.75)0.58 (0.47–0.71)
 Studies that required Barrett's  esophagus for EAC diagnosis53.770.4400.44 (0.33–0.58)0.43 (0.33–0.56)
 Adjusted results713.250.04550.50 (0.34–0.74)0.52 (0.41–0.66)
H. pylori (the broader definition)1316.600.17280.56 (0.45–0.69)0.57 (0.46–0.71)
CagA-negative strains51.640.8001.08 (0.76–1.53)1.07 (0.75–1.52)
CagA-positive strains**54.780.31160.41 (0.28–0.62)0.40 (0.28–0.56)
  • 2 Q statistic for homogeneity in random effects model.

  • P value for the Q statistic in random effects model.

  • Higgins I2 statistic for heterogeneity in random-effects model.

  • §Being positive for antibodies against whole-cell or being positive for H. pylori in invasive tests; numbers inside parenthesis refer to number of cases/controls.

  • Being positive for antibodies against whole-cell or CagA, or positive for H. pylori in invasive tests.

  • This row compares subjects with CagA-negative strains versus those who are H. pylori negative.

  • **This row compares subjects with CagA-positive strains versus those who are H. pylori negative.