TABLE 3.
Assessment of directness (0–1 points)1
| 0 points (indirectness) | 1 point |
| 1) Differences in population: award 0 points only if there are important reasons to think that the physiology of the population of interest does not conform with that of the population tested, potentially leading to a considerably different effect measure | No important differences in the population or intervention; hard clinical outcome |
| 2) Differences in intervention: applicable in RCTs; e.g., supervised vs. nonsupervised exercise; systematic difference in care between intervention and control group | |
| 3) Surrogate markers (e.g., blood lipids, blood pressure) | |
| 4) Network meta-analyses of RCTs; only a few available to date in nutrition research |
1
RCT, randomized controlled trial.