Table 1.
| Importance of a potential determinanta | |||
| association across primary samples | |||
| % | direction | ||
| ++ | 100 | positive or negative | |
| + | >75 | positive or negative | |
| 0 | ≤75 | positive or negative and | |
| ≤75 | no association | ||
| – | >75 | no association | |
| –– | 100 | no association | |
| Strength of the evidenceb | |||
| ‘sufficient evidence’ | ‘consistency’ | ||
| reviews | independent cohorts | across primary samples | |
| n | n | % | |
| Ce | ≥3 | ≥2 | 100 |
| Pe | ≥2 | ≥2 | >75 |
| Ls | ≥1 | ≥1 | >66 |
| Lnc | ≥1 | 0 | ≤66 |
Ce Convincing evidence, Pe Probable evidence, Ls Limited, suggestive evidence, Lnc Limited, non-conclusive evidence
aImportance was evaluated based on the proportion of study that reported a positive or negative association between a potential determinant and PA. E.g., a potential determinant was scored ‘++’ if 100% of eligible samples reported either a positive or a negative association with PA
bStrength of the evidence was evaluated based on the number of reviews, the reported study design of eligible primary studies, and the consistency across primary samples. For each level of evidence, each criterion for number of reviews, study design, and consistency had to be fulfilled. E.g., there was ‘convincing evidence’ (Ce, highest level of evidence), if the results were: (1) based on a substantial number of reviews (here defined as ≥3 SLRs, [70]) including data of different study designs and (2) based on at least two independent primary cohort studies and, (3) showed a consistent association with PA (here defined as 100% of eligible samples reported associations to be in the same direction)