Skip to main content
. 2022 Aug 4;13:921689. doi: 10.3389/fpsyt.2022.921689

TABLE 3.

Moderated mediation regression analysis for repeated measure designs (n = 66).

PANSS depressed factor Conditional total, direct and indirect effects Indices of moderation by group


Effects Moderator (Group) Mediator Average (VO2max) Effect SE LLCI ULCI p Effect SE LLCI ULCI P
Total effects AVG 0.16 0.39 −0.62 0.95 0.682 −1.23 0.59 −2.42 0.05 0.042
HIIT −1.07 0.45 −1.96 −0.18 0.019
Direct effects AVG −0.26 0.13 0.36 −0.59 0.85 0.721 −0.84 0.56 −1.95 0.27 0.138
HIIT 0.32 −0.70 0.42 −1.55 0.15 0.103
Indirect effects AVG 0.03 0.09 −0.16 0.22 −0.40 0.32 −1.19 0.09
HIIT −0.37 0.31 −1.15 0.07

Moderated mediation analysis: Outcome (Y) is the difference in depressive symptoms (post-pre-intervention PANSS depressed factor score), moderator (W) is the group (AVG = 0/HIIT = 1), and mediator (M) is the difference in VO2max (post-pre-intervention). Effect, Regression coefficient; SE, standard error; LLCI-ULCI, 95% Confidence interval; Total effect, The total conditional effect; Direct effect, The conditional effect not mediated by VO2max; Indirect effect, The conditional effect through VO2max.

For indirect effects, the standard errors and confidence intervals are bootstrapped.