Table 3.
Subgroup analyses of summary correlation between PA and MPA. Pa values for the within-subgroup effect sizes were calculated with the z test; Pb values for between-subgroup differences were calculated with the Q test; and Pc values for heterogeneity within subgroups were calculated with the Q test.
| Moderator | Studies, n | Summary r (95% CI) | Pa value | Heterogeneity | Pb value | ||
|
|
I2 (%) | Pc value |
|
||||
| Time of data collection | .14 | ||||||
|
|
Before COVID-19 | 4 | –0.333 (–0.466 to –0.187) | <.001 | 97.555 | <.001 |
|
|
|
During COVID-19 | 13 | –0.207 (–0.285 to –0.126) | <.001 | 96.438 | <.001 |
|
| Country or region | .71 | ||||||
|
|
Developed regions | 2 | –0.446 (–0.616 to 0.236) | .39 | 99.133 | <.001 |
|
|
|
China | 7 | –0.201 (–0.311 to –0.127) | <.001 | 97.873 | <.001 |
|
|
|
Other developing regions | 8 | –0.217 (–0.326 to –0.103) | <.001 | 63.310 | <.001 |
|
| Population | .26 | ||||||
|
|
Young adults | 15 | –0.250 (–0.325 to –0.173) | <.001 | 96.681 | <.001 |
|
|
|
Adolescents | 2 | –0.129 (–0.333 to 0.086) | .24 | 97.787 | <.001 |
|