Recreational Football for Young-to-middle-aged Inactive Men

Randomization:The participants were randomized 1:2 to the control or the intervention groups,respectively. Randomization was performed by a researcher not involved in data collectionor outcome assessment. Blinding of participants and instructors, coaches and theinvestigation team leading the training sessions was not feasible in this type ofexercise intervention. Outcome evaluators, who were not involved in intervention deliveryor group allocation, were blinded to group assignment throughout the assessment timepoints (baseline and 12 weeks).Statistical analysis:Both intention‐to‐treat (ITT) and per‐protocol (PP) approaches were applied. The ITTanalysis included all randomized participants; missing post‐intervention values werehandled using multiple imputation. The PP analysis included only participants whoattended ≥66% of the prescribed sessions. Descriptive statistics were calculated usingobserved data and are reported as mean ± SD. The distribution of continuous variables wasinspected visually using histograms and Q‐Q plots, and normality was further assessedusing the Shapiro‐Wilk test. Model assumptions were evaluated by visual inspection ofresidual plots and Q‐Q plots of model residuals.Outcome variables were analysed using linear mixed‐effects models to account for repeatedmeasurements within participants. For each outcome, Time (pre vs post), Group(intervention vs control), and the Time × Group interaction were specified as fixedeffects, with a participant‐specific random intercept to model within‐subject clustering,according to the general form: outcome Time + Group + Time × Group + (1 | participant).Fixed effects were evaluated using omnibus F‐tests with Satterthwaite‐approximateddegrees of freedom, and model estimates are reported as regression coefficients (b) withstandard errors and 95% confidence intervals. Where appropriate, post hoc pairwisecomparisons of estimated marginal means were used to decompose significant effects, withmultiplicity controlled using Bonferroni adjustment. Model fit was summarised usingmarginal and conditional R², and variance components were reported as random‐interceptand residual variance. The intraclass correlation coefficient (ICC) was calculated toquantify the proportion of variance attributable to between‐participant differences.Effect sizes were expressed as partial eta squared (ηp²) for omnibus effects and Cohen'sd for within‐ and between‐group contrasts, using the standard deviation (SD), not thestandard error, as the standardiser to avoid inflation (Lakens, 2013). Cohen's d wasinterpreted as trivial (<0.20), small (0.20‐0.49), moderate (0.50‐0.79), and large(≥0.80), following Cohen's conventional thresholds (Cohen, 1988). Partial eta squared wasinterpreted as small (ηp² ≈ 0.01), moderate (ηp² ≈ 0.06), and large (ηp² ≈ 0.14)(Richardson, 2011). These thresholds were used only as descriptive guidelines, andeffects were interpreted alongside the magnitude and precision of the estimated meandifferences. All tests were two‐sided, with statistical significance set at α = 0.05.Statistical analyses were performed using jamovi (version 2.6.23; The jamovi project,Sydney, Australia) and R packages (R Core Team, R Foundation for Statistical Computing,Vienna, Austria) appropriate for each analysis.To overcome the common limitation of low statistical power highlighted in previousrecreational football interventions, a substantially larger sample size was deliberatelyrecruited. This strategic design choice was undertaken not only to enhance the robustnessof subgroup and ITT analyses but also to increase the precision of effect size estimates,thereby improving the generalizability of the findings.