Relationships between walking speed, activities and participation in people with chronic stroke in Burundi.

The cross-sectional design measured people at one time, so it cannot show that changing walking speed changed balance, daily activity or social participation. Participants were recruited purposively from hospital and rehabilitation centre records, not randomly, so they may not represent all stroke survivors in Burundi. Only people able to walk 10 metres with or without an aid were included, so the study says little about stroke survivors who cannot walk. Major cognitive impairment and other permanent neurological disease were excluded, so findings may not apply to the whole stroke population. The three walking-speed groups had unequal numbers, which could have affected the statistical comparisons. Activity and participation scores were patient-reported, which can introduce recall bias. Perry walking-speed categories were developed in a high-income context and may not fit the local environment and rehabilitation access in Burundi.

Do not assume that making people walk faster will improve their social participation. This study measured people at one time, and walking speed was not related to social participation overall.