PMID 40753404

Research
All papers

High-intensity interval training with robot-assisted gait therapy vs. treadmill gait therapy in chronic stroke: a randomized controlled trial

Kim J, Do J, Bae CR, Mo YH, Kim JH, Kim DY

Question

Does combining HIIT with robot-assisted gait therapy improve gait, balance, lower-limb function, cardiopulmonary function, activities of daily living, and body composition more than treadmill-based gait therapy in patients with chronic stroke?

Summary

This exploratory randomized controlled trial compared HIIT integrated into robot-assisted gait therapy with treadmill-based gait therapy in adults with chronic stroke and dependent ambulation. Both groups trained for 30 minutes, three times weekly for 8 weeks. The robot-HIIT protocol alternated 2 minutes at moderate perceived exertion with 3 minutes at high perceived exertion, using an end-effector robot and continuous heart-rate monitoring. Forty-four of 48 randomized participants completed outcome assessments. The robot-HIIT group improved more than treadmill therapy on walking speed, ambulation category, balance, walking endurance, and lower-limb motor function, while VO2max improved within the robot-HIIT group but did not differ significantly between groups. The findings are promising for stroke rehabilitation but are not directly generalizable to unsupervised consumer HIIT.

Methodology

  • Adults with chronic hemorrhagic or ischemic stroke, gait disturbance, FAC score 3 or less, and more than 6 months post-stroke.
  • 48 participants.
  • End-effector robot-assisted gait therapy using Morning Walk.
  • Work intervals: 3 minutes high intensity.
  • Recovery: 2 minutes moderate intensity.
  • Intensity: Moderate RPE 10-12 and high RPE > 14; target heart rate calculated using heart-rate reserve.
  • 30 minutes.
  • 3 sessions/week.
  • 8 weeks.
  • Prospective randomized controlled trial without blinding.
  • 10-meter walk test, Functional Ambulation Category, Berg Balance Scale, and 2-minute walk test were tracked.

Outcomes

10-meter walk test

The primary outcome improved significantly more in robot-HIIT than treadmill control; robot-HIIT changed from 0.41 +/- 0.23 to 0.55 +/- 0.29 m/s while control changed from 0.50 +/- 0.22 to 0.52 +/- 0.24 m/s.

Between-group change 0.14 +/- 0.13 vs 0.01 +/- 0.08; p < 0.001; d = 1.2; CI [0.065, 0.198].

Improved

Functional Ambulation Category

Both groups improved, with greater improvement in robot-HIIT.

Control 2.72 +/- 0.63 to 3.59 +/- 1.00; robot-HIIT 2.40 +/- 0.73 to 3.81 +/- 1.36; between-group p = 0.009; d = 0.79 reported for change comparison.

Improved

Balance and lower-limb motor function

BBS and FMA-LE improved in both groups and improved more in robot-HIIT.

BBS between-group p = 0.012 in table and p = 0.015 in abstract; FMA-LE between-group p < 0.001; FMA-LE change d = 1.27.

Improved

2-minute walk test

Only robot-HIIT showed a significant within-group improvement in 2MWT, and between-group difference favored robot-HIIT.

Control 68.22 +/- 32.29 to 72.59 +/- 34.16 m; robot-HIIT 56.81 +/- 38.25 to 74.50 +/- 42.27 m; between-group p = 0.008 in table and p = 0.005 in abstract.

Improved

VO2max

VO2max improved significantly within robot-HIIT, but the between-group difference was not significant.

Robot-HIIT 13.50 +/- 9.31 to 20.74 +/- 9.03 ml/kg/min; control 21.59 +/- 10.26 to 25.69 +/- 14.76 ml/kg/min; between-group p = 0.901.

Mixed

Modified Barthel Index and lean body mass

MBI improved within the control group only and did not differ significantly between groups; lean body mass did not significantly change or differ between groups.

MBI between-group p = 0.473; lean body mass between-group p = 0.613.

Mixed

Insights

  • In clinical populations with neurological impairment, RPE may be necessary alongside heart-rate targets because medications, spasticity, weakness, and fatigue can distort heart-rate based intensity monitoring.
  • Standard interval durations can be paired with individualized cadence and modality selection to hit intensity targets.
  • Clinical HIIT findings involving robotics and therapist supervision should not be generalized to consumer bodyweight HIIT without strong caveats.
  • For chronic stroke rehabilitation, combining task-specific gait practice with high-intensity intervals may be more promising than treadmill gait therapy alone.

Limitations

  • Small exploratory sample.
  • No blinding and potential observer bias.
  • Single tertiary hospital in South Korea.
  • Chronic stroke-only sample limits generalization to acute or subacute stroke.
  • Baseline lower-limb function differed between groups because allocation was not stratified.
  • Mechanisms were not directly studied.
  • No RAGT-alone or HIIT-alone group, so independent effects cannot be isolated.
  • Only one standardized HIIT protocol was tested.
  • Adverse-event reporting was not explicit in the extracted main text.

Safety

  • The extracted main text did not report explicit adverse-event counts or injury events.
  • Patients with uncontrolled cardiovascular or other physical conditions preventing high-intensity gait training were excluded.
  • Continuous heart-rate monitoring, therapist presence, RPE adjustment, and robot body-weight support were used during intervention sessions.