TABLE 2.
Risk-of-bias assessment by Cochrane RoB tool1
| Overall risk | |||||||
|---|---|---|---|---|---|---|---|
| First author, year (reference) | 1. Randomization | 2. Intervention | 3. Missing data | 4. Measurement data | 5. Selection results | Cochrane | Adjusted |
| Protein | |||||||
| Juncos, 1987 (48) | High2 | Low | Low | Low | Low | High | Low |
| Pincus, 1987 (49) | High2 | Low | Low | Low | Low | High | Low |
| Frankel, 1989 (42) | High2 | Moderate | Low | Low | Low | High | Moderate |
| Berry, 1991 (43) | Low | Low | Low | Low | Moderate | Moderate | Moderate |
| Bracco, 1991 (50) | High2 | Low | Low | Low | Low | High | Low |
| Karstaedt, 1992 (46) | High2 | Moderate | Low | Low | Low | High | Moderate |
| Karstaedt, 1993 (51) | High2 | Low | Low | Low | Low | High | Low |
| Simon, 2004 (39) | High | High | Low | Low | Low | High | High |
| Barichella, 2006 (52) | Low | Low | Low | Low | Low | Low | Low |
| Cucca, 2015 (44) | Low | Low | Moderate | Low | Low | Moderate | Moderate |
| Vitamins | |||||||
| Lamberti, 2005 (53) | High2 | Low | Low | Low | Low | High | Low |
| Postuma, 2006 (47) | Low | Low | Low | Low | Moderate | Moderate | Moderate |
| Lee, 2010 (28) | Low | Low | Low | Low | Low | Low | Low |
| Müller, 2013 (54) | High2 | Low | Low | Low | Low | High | Low |
| Rispoli, 2017 (40) | High2 | High | Low | Low | Low | High | High |
| Habibi, 2018 (41) | Low | Low | High | Moderate | High | High | High |
| Nagayama, 2004 (55) | High2 | Low | Low | Low | Low | High | Low |
| Fiber | |||||||
| Astarloa, 1992 (56) | High2 | Low | Low | Low | Low | High | Low |
| Fernandez, 2014 (57) | Low | Low | Low | Low | Low | Low | Low |
| Caffeine | |||||||
| Deleu, 2006 (58) | Low | Low | Low | Low | Low | Low | Low |
| Soybeans | |||||||
| Nagashima, 2016 (59) | High2 | Low | Low | Low | Low | High | Low |
| Ketogenic diet | |||||||
| Elbarbry, 2019 (45) | High2 | Low | Low | Low | Moderate | High | Moderate |
1The RoB, as measured with the Cochrane tool, is assessed for 5 individual domains: 1) randomization process, 2) deviations from intended interventions, 3) missing outcome data, 4) measurement of outcome data, and 5) selection of reported results. The RoB is defined as low, moderate, or high and is inversely related to the overall quality. Following the official Cochrane algorithm, the overall RoB is rated low (all individual domains have low RoB), moderate (moderate RoB at least in 1 domain), or high RoB (at least 1 domain high RoB or multiple essential domains moderate RoB). Nonrandomized trials automatically score a high RoB (indicated with a superscript “2”) for the randomization process with the Cochrane tool. Therefore, a high RoB in domain 1 is not taken into account for the overall adjusted RoB judgment of non-randomized studies. RoB, risk of bias.
2Nonrandomized trial.