The primary objective of the study is to investigate the energy consumption, cardiorespiratory load and perceived exertion, and how these parameters change, during walking with robot-assistance compared to walking on a treadmill and walking overground in stroke patients.
A secondary objective is to investigate whether these changes or differences in energy consumption, cardiorespiratory load and perceived exertion during walking with and without robot-assistance in stroke patients are related to changes or differences spatiotemporal gait characteristics.
Background. Impaired cardiorespiratory fitness, which is a major risk factor in the development of cardiorespiratory diseases, is frequently reported in stroke patients. The mean energy cost of walking, i.e. the amount of oxygen consumption in milliliter per kilogram of body-weight per meter, in stroke patients is almost twice as high compared to healthy subjects (resp. 0.27 ml/kg/m vs. 0.15 ml/kg/m). In the rehabilitation of stroke patients, the primary aim is to improve kinematic and functional gait-related parameters. However, due to the previously mentioned cardiorespiratory risks, it is important to be aware of the energy consumption and cardiorespiratory load of stroke patients during gait rehabilitation. In the past, gait training was mainly fulfilled by treadmill training, overground training and/or more conventional therapies, but in recent years, the implementation of robot-assistance in gait rehabilitation is increasing. However, what the influence is of robot-assistance on the cardiorespiratory load and energy consumption, and therefore also what potentially negative and/or positive side effects are for the cardiorespiratory system, is less investigated and unclear.
Up to now, short walking durations of robot-assisted gait (up to 7 minutes) seem less energy consuming and cardiorespiratory stressful than walking without robot-assistance. However, what the influences are of longer walking durations is not clear. In addition, it is also unclear why possible differences between robot-assisted gait and walking without robot-assistance might exist. One possible explanation might be that differences in spatiotemporal gait parameters are responsible for differences in energy consumption and cardiorespiratory load.
Patient recruitment. Stroke patients in the Rehabilitation Centre St. Ursula (Herk-de-Stad, Belgium) will receive verbal and written information on the aims and interventions of the study. Eligible stroke patients, who agree to participate in the study, will be recruited. Signed informed consent will be obtained from all participants.
Sample size. Sample size calculation is based on previous investigations indicating large effect sizes between the effect of robot-assisted gait compared to walking without robot-assistance on energy consumption and cardiorespiratory load (based on a systematic review submitted for peer-review). To detect a large effect size (f = 0.40) of robot-assisted gait compared to overground and treadmill gait on energy consumption, cardiorespiratory load and perceived fatigue, in a repeated measures within subjects design (3 walking conditions and 4 measurements), with a significance level of 5% and a power level of 80%, a sample size of 21 subjects is needed (G*Power 3.1 for Mac). Sample size is inflated up to 24 subjects, so each walking order will be performed the same number of times.
Intervention. Patients will be tested in 3 single walking sessions each on a separate day: walking in the Lokomat with 60% guidance force, walking on a treadmill and walking overground. Within subjects, all walking conditions will be performed at the same comfortable walking speed (CWS), with the same amount of body-weight support (BWS) (if necessary) during a total duration of maximum 30 minutes. The CWS (with a maximum of 3.2 kmph corresponding to the maximum Lokomat speed) and the amount of BWS (if necessary) will be individually determined on a separate day before the start of the study. Walking tests will be terminated early when relative or absolute indications are presented as reported by the American Heart Association or when patients are unable to continue walking. Patients will be asked to not consume food, alcohol, caffeine or nicotine at least 3 hours prior to the intervention, and not to perform additional strenuous activities at least 12 hours prior to the interventions. Walking sessions will be controlled for time of day. Before the start of the study, demographic and clinical characteristics will be collected and the CWS and the amount of BWS (if necessary) will be determined in a 10 minute walking test. At the start of each walking condition, a chest-carrying gas analysis system with mouth mask (Metamax 3B, Cortex, Germany), a heart rate belt (Polar H7) and 2 wearable foot sensors (Physiolog, Gait Up, Switzerland) will be applied. Patients will be seated for 5 minutes during which resting values (energy consumption, cardiorespiratory parameters and perceived fatigue) will be registered. After a resting period of 5 minutes, patients will walk for 30 minutes during which energy consumption, cardiorespiratory parameters, perceived fatigue and spatiotemporal parameters will be monitored continuously. Perceived fatigue will be registered every minute. Average values at rest, the beginning, middle and end of the walking sessions will be calculated offline.
Randomization and Concealment. Walking sessions will be performed in a random order at 3 separate days. An independent investigator will assign the 24 patients (in 2 series of 12) at random to one of the 6 possible walking orders using a random sequence generator. Allocation will be concealed for the investigators using an excel file with blind and locked sections, to which only the independent investigator has access to. The random walking order of the patient will therefore only be available when the patient has been recruited and his name is entered in the excel sheet. This method will assure that the investigator does not know the walking order of the next participant.
Dropout. In case subjects drop out, the subject will be replaced by a new participant who will perform all three trials in the same randomized order as the subject that dropped out. So, in case of drop out, additional patients will be tested until the data of 24 patients that participated in all three conditions are collected.
Statistical analysis. Statistics will be performed using SPSS (IBM, Chicago, IL). Descriptive statistics will be calculated for baseline demographic and clinical patient characteristics. Repeated measures analyses of variance (ANOVA) with Bonferroni correction for multiple comparisons will be used to analyze differences in primary and secondary outcomes within and between walking conditions. Regression analysis will be performed to evaluate whether (changes in) spatiotemporal parameters are predictive for (changes in) energy consumption. The significance level will be set at 5%.
Inclusion Criteria:
Exclusion Criteria: