PURPOSE: Locomotor high-intensity interval training (HIIT) has been shown to improve walking capacity (speed and endurance) more than moderate-intensity aerobic training (MAT) after stroke, but it is unclear which training parameter(s) should be prioritized (e.g. speed, heart rate, blood lactate, step count) and to what extent walking capacity gains are the result of gait changes versus cardiorespiratory adaptations. This study aimed to assess which training parameters and longitudinal adaptations most strongly mediate 6-minute walk distance (6MWD) gains from post-stroke HIIT. METHODS: The HIT-Stroke Trial randomized 55 persons with chronic stroke and walking limitations to HIIT or MAT
each involving overground and treadmill walking, 45 minutes, 3 times/week for 12 weeks. HIIT used short maximum speed intervals, targeting >
60% heart rate reserve (HRR). MAT targeted 40-60% HRR. Blinded outcomes included 6MWD, short-distance gait function (e.g. fastest 10-meter gait speed) and aerobic capacity (e.g. ventilatory threshold). This ancillary analysis used structural equation models to compare mediating effects of different training parameters and longitudinal adaptations on 6MWD changes. RESULTS: Net gains in 6MWD from HIIT versus MAT were primarily mediated by faster training speeds (+57 meters, p <
.0002) and longitudinal gait adaptations (+37 meters, p = .0005). Training step count was also positively associated with 6MWD gains, but was lower with HIIT versus MAT, which decreased the net 6MWD gain from HIIT (-14 meters, p = .02). HIIT generated higher training heart rate and lactate than MAT, but aerobic capacity gains were similar between groups, and 6MWD changes were not associated with training heart rate, training lactate, or aerobic adaptations. CONCLUSIONS: To increase walking capacity with post-stroke HIIT, training speed and step count appear to be the most important parameters to prioritize.