TABLE 1.
Summary information on characteristics of studies contributing to main analysis
| n/median (%/IQR) | |
|---|---|
| N total trials | 169 |
| Sample size, median (IQR) | 126 (65‐251) |
| Age (y), median (IQR) | 49 (45‐56) |
| % Female, median (IQR) | 73 (51‐100) |
| Mean BMI (kg/m2), median (IQR) | 34 (31‐36) |
| Number of arms contributing to analysis, n studies (%) | |
| 2 | 136 (80) |
| 3 | 27 (16) |
| 4 | 4 (2) |
| 6 | 1 (1) |
| 7 | 1 (1) |
| Geographical region, n (%) | |
| North America | 94 (56) |
| South America | 2 (1) |
| Europe | 52 (31) |
| Asia | 7 (4) |
| Australia and New Zealand | 0 (0) |
| Africa | 12 (7) |
| Mixed (Australia and Europe) | 1 (1) |
| Missing | 1 (1) |
| Selection based on disease status, n (%) | |
| No | 98 (58) |
| Yes | 70 (41) |
| Missing | 1 (1) |
| Recruitment method, n (%) | |
| Self‐initiated | 67 (40) |
| Prompted | 72 (43) |
| Missing | 30 (18) |
| Random sequence generation (selection bias), n (%) | |
| Low | 101 (60) |
| Unclear | 67 (40) |
| High | 1 (1) |
| Allocation concealment (selection bias), n (%) | |
| Low | 63 (37) |
| Unclear | 106 (63) |
| High | 0 (0) |
| Blinding of outcome (detection bias), n (%) | |
| Low | 147 (87) |
| Unclear | 14 (8) |
| High | 8 (5) |
| Incomplete outcome data (attrition bias), n (%) | 169 (100) |
| Low | 141 (83) |
| Unclear | 7 (4) |
| High | 21 (12) |
| Other bias, n (%) | |
| Low | 2 (1) |
| Unclear | 13 (8) |
| High | 17 (10) |