"Magic" Number of Treadmill Sessions Needed to Achieve Meaningful Change in Gait Speed After Stroke: A Systematic Review.

Only 6 of 17 included studies were rated good to excellent on the PEDro scale; the majority were fair or poor quality. The review protocol was not registered prospectively (PROSPERO requirements changed during data collection). Single-session studies were included, which the authors acknowledge may have skewed the data. All adjuncts (body-weight support, FES, brain stimulation, exoskeletons, mirror therapy) were excluded, so the findings apply only to plain treadmill walking. Most participants could already walk independently; the results do not extend to patients who cannot walk on a treadmill without support. Only chronic stroke (onset >6 months) was included; nothing here applies to the acute or subacute recovery window. Several included studies had very small sample sizes, reducing their power. Wide standard deviations in session counts across groups make it hard to draw firm conclusions from the averages. The analysis relied on group-level averages, not individual patient trajectories, so individual variability is not captured.

Funded by NIH extramural research support. The authors state the funders played no role in the design, conduct, or reporting of the study.

Do not treat '30 sessions' as a validated prescription. The correlation between session count and gait speed change was not statistically significant (p = 0.17), and the authors explicitly state that no firm session number can be recommended from this data. The parameter that was significant was minutes per session, not total sessions. A group that completed 78 sessions still did not reach the MCID, so more visits do not guarantee more gain.