Twenty-one RCTs (n = 904) were included in the review. The median PEDro score was 7 out of a maximum 11 (range 4 to 8).
Gait speed (17 RCTs, n = 692). A pooled analysis of 12 RCTs of gait-oriented training showed a significant effect of intervention (summarised effect size 0.45, 95% CI: 0.27 to 0.63, p<0.01).
Walking distance (13 RCTs, n = 743). A pooled analysis of nine RCTs of gait-oriented training showed a significant effect of intervention (summarised effect size 0.62, 95% CI: 0.30 to 0.95, p<0.01). There was no significant effect in studies which targeted lower limb strengthening (three RCTs).
Significant heterogeneity was found for a number of the analyses.
Results of a best-evidence synthesis were also reported; no significant differences were reported in studies of limb strengthening (two RCTs), improved stair climbing ability was found with cardiorespiratory training (one RCT) and improved functional ambulation with gait-oriented training (two RCTs). A number of non-significant results were also reported for this last group of trials.