(b) Subgroup comparisons for Helicobacter pylori infection on the risk of esophageal adenocarcinoma
| Subgroup | No. of studies | I 2 (%) | Overall OR (95% CI) | P value§ | |
|---|---|---|---|---|---|
| DL | HKSJ | ||||
| Esophageal adenocarcinoma | |||||
| Case/control (7687/301085) | |||||
| All studies | 20 | 73.4 | 0.55 (0.43, 0.70) | 0.23 (0.15, 0.36) | |
| Ethnicity | <0.001∗ | ||||
| Caucasians | 18 | 33.9 | 0.60 (0.50, 0.73) | 0.23 (0.14, 0.37) | |
| Mixed race | 1 | — | 0.29 (0.25, 0.35) | — | |
| Australian | 1 | — | 0.52 (0.36, 0.76) | — | |
| Study area | 0.100 | ||||
| North America | 11 | 80.5 | 0.62 (0.42, 0.93) | 0.54 (0.35, 0.81) | |
| Europe | 7 | 56.2 | 0.49 (0.34, 0.69) | 0.13 (0.06, 0.27) | |
| Middle East | 1 | — | 0.32 (0.10, 1.02) | — | |
| Oceania | 1 | — | 0.52 (0.36, 0.76) | — | |
| Study type | 0.287 | ||||
| Case-control | 19 | 74.4 | 0.56 (0.44, 0.72) | 0.23 (0.14, 0.37) | |
| Cohort | 1 | — | 0.23 (0.06, 0.83) | — | |
| Study object | <0.001∗ | ||||
| Population-based studies | 11 | 37.9 | 0.59 (0.48, 0.73) | 0.15 (0.08, 0.28) | |
| Clinical-based studies | 9 | 73.4 | 0.52 (0.35, 0.79) | 0.47 (0.29, 0.77) | |
| Matching | <0.001∗ | ||||
| Studies with matched controls | 5 | 0 | 0.73 (0.57, 0.92) | 0.71 (0.60, 0.84) | |
| Studies without matched controls | 15 | 73.3 | 0.48 (0.36, 0.65) | 0.21 (0.13, 0.36) | |
| Helicobacter pylori detection method | <0.001∗ | ||||
| Serology | 14 | 36.3 | 0.62 (0.51, 0.75) | 0.19 (0.11, 0.32) | |
| Histology | 5 | 62.1 | 0.39 (0.26, 0.60) | 0.39 (0.18, 0.84) | |
| Serology, histology | 1 | — | 0.81 (0.32, 2.07) | — | |
Note: DL: DerSimonian-Laird; HKSJ: Hartung-Knapp-Sidik-Jonkman. §Differences between subgroups; ∗the difference was statistically significant.