Participants: The mean age of participants across the studies ranged from 50 to 74 years. The mean time after stroke ranged from 1.6 to 27 months, and one study did not report this information. Participants were recruited from people living in the community in 55% of the trials. Intervention: In all studies, the experimental group received treadmill training without body weight support. Participants undertook training for 25 to 40 min, 3–5/wk,

for 2.5 to 26wk. The control group received no intervention (three studies), a Libraries non-walking intervention (four studies), or overground walking (three studies). Outcome measures: Walking speed was measured selleck kinase inhibitor using the 10-m Walk Test (eight studies) and results were converted to m/s. Walking distance was measured using the 6-min Walk Test (seven studies) and results were converted to m. Walking speed: The immediate effect of treadmill training versus no intervention or a non-walking intervention on walking speed was examined by pooling data from seven studies ( Ada et al 2003, Eich et al 2004, Weng et al 2006, Ivey et al 2011, Kuys et al 2011, Olawale et al 2011, Ada et al 2013) involving 275 participants. Treadmill training increased walking speed 0.14 m/s (95% CI 0.09 to 0.19) more than no intervention/non-walking intervention ( Figure 2a, see Figure 3a on the eAddenda for the detailed forest plot). The effect of treadmill

training beyond the intervention selleck period compared with no intervention/non-walking intervention on walking speed was examined by pooling data from four studies ( Ada et al 2003, Eich et al 2004,

Kuys et Thiamine-diphosphate kinase al 2011, Ada et al 2013) involving 167 participants. Treadmill training increased walking speed 0.12 m/s (95% CI 0.08 to 0.17) more than no intervention/ non-walking intervention ( Figure 2b, see Figure 3b on the eAddenda for the detailed forest plot). The immediate effect of treadmill versus overground training on walking speed was examined by pooling data from three studies (Pohl et al 2002, Langhammer and Stanghelle 2010, Olawale et al 2011) involving 119 participants. There was no significant difference in walking speed between treadmill training and overground training (MD 0.05 m/s, 95% CI −0.12 to 0.21) (Figure 4, see Figure 5 on the eAddenda for a detailed forest plot). No studies measured the effect of treadmill training versus overground walking on walking speed beyond the intervention period. Walking distance: The immediate effect of treadmill training versus no intervention or a non-walking intervention on walking distance was examined by pooling data from six studies ( Ada et al 2003, Eich et al 2004, Ivey et al 2011, Kuys et al 2011, Olawale et al 2011, Ada et al 2013) involving 249 participants. Treadmill training increased walking distance 40 m (95% CI 27 to 53) more than no intervention/non-walking intervention ( Figure 6a, see Figure 7a on the eAddenda for the detailed forest plot).