Sage Journals: Discover world-class research

Abstract

Introduction. Physiological responses are rarely considered during walking after stroke and if considered, only during a short period (3-6 minutes). The aims of this study were to examine physiological responses during 30-minute robot-assisted and body weight–supported treadmill and overground walking and compare intensities with exercise guidelines. Methods. A total of 14 ambulatory stroke survivors (age: 61 ± 9 years; time after stroke: 2.8 ± 2.8 months) participated in 3 separate randomized walking trials. Patients walked overground, on a treadmill, and in the Lokomat (60% robotic guidance) for 30 minutes at matched speeds (2.0 ± 0.5 km/h) and matched levels of body weight support (BWS; 41% ± 16%). Breath-by-breath gas analysis, heart rate, and perceived exertion were assessed continuously. Results. Net oxygen consumption, net carbon dioxide production, net heart rate, and net minute ventilation were about half as high during robot-assisted gait as during body weight–supported treadmill and overground walking (P < .05). Net minute ventilation, net breathing frequency, and net perceived exertion significantly increased between 6 and 30 minutes (respectively, 1.8 L/min, 2 breaths/min, and 3.8 units). During Lokomat walking, exercise intensity was significantly below exercise recommendations; during body weight–supported overground and treadmill walking, minimum thresholds were reached (except for percentage of heart rate reserve during treadmill walking). Conclusion. In ambulatory stroke survivors, the oxygen and cardiorespiratory demand during robot-assisted gait at constant workload are considerably lower than during overground and treadmill walking at matched speeds and levels of body weight support. Future studies should examine how robotic devices can be Future studies should examine how robotic devices can be exploited to induce aerobic exercise.

Keywords

energy metabolism aerobic exercise walking exoskeleton devices body weight support stroke

Introduction

After stroke, cardiorespiratory fitness is low—about half that of healthy individuals^1-3—increasing the risk of developing secondary cardiovascular diseases.⁴ Two-thirds of patients with stroke subsequently die from cardiovascular diseases, making cardiovascular diseases the leading cause of death (4 times higher rate than the general population).⁵ This high rate stresses the importance of implementing fitness training in stroke rehabilitation.

Aerobic exercise improves cardiorespiratory fitness after stroke.^6,7 Common forms of aerobic exercise are walking (overground or on a treadmill) and cycling (arm or legs).⁸ Stroke guidelines recommend large-muscle exercising activities 3 to 5 times per week for 20 to 60 minutes at moderate intensity.⁹ The American Stroke Association defines moderate intensity as 40% to 70% of peak oxygen consumption, 40% to 70% of heart rate reserve, 55% to 80% of maximum heart rate, and 11 to 14 rating of perceived exertion.⁹

Patients with stroke spend the majority of rehabilitation time on practicing walking (25% to 45%).¹⁰ Hence, walking offers a time-efficient way of fulfilling 2 important rehabilitation goals at once: improving gait and improving cardiorespiratory fitness. Because physiotherapists’ time is sparse—patients often receive less than the recommended minimum of 45 minutes to 1 hour of physiotherapy per weekday^11-15—it is important that physiotherapists use their time highly efficiently. To improve gait, daily practice is desirable; to improve cardiorespiratory fitness, a minimum of 3 times per week is recommended.⁹

Current inpatient stroke rehabilitation induces no or only little cardiovascular stress to result in a cardiorespiratory training effect.^16,17 Patients’ physical impairment and lack of therapeutic staff hinder the implementation of fitness training in inpatient rehabilitation.¹⁸ The use of recumbent steppers and cycle ergometers, or body weight support (BWS) systems and robotic devices may solve these barriers. Especially these latter two facilitate the administration of aerobic exercise to those who cannot walk in a task-specific manner while reducing the number of required therapists.¹⁹ However, physiological responses during robot-assisted gait are rarely considered,²⁰ making it unclear what the exercise intensity of this type of training is and whether it elicits adequate intensity to reach an aerobic training effect. The Lokomat is an example of a robotic device used in gait rehabilitation after stroke. This device automates locomotion on a treadmill by providing robotic guidance at the knee and hip joints. Varying degrees of robotic guidance (0%-100%), BWS (0%-100%), and walking speed (0-3.2 km/h) can be set according to the patient’s abilities.²¹ These numerous combinations of settings lead to different effects on individual exercise intensity. Although decreasing the body weight and assisting the hip and knee joints may allow longer durations of task-oriented practice, it may also decrease exercise intensity below recommended levels for aerobic exercise. In turn, robotic assistance may allow walking faster, which may increase exercise intensity.

Current studies—which are few in number—indicate that the oxygen consumption and cardiorespiratory load during robot-assisted gait after stroke is low (lower than walking without robot assistance).²⁰ van Nunen et al²² classified Lokomat walking as light-intensity exercise in 10 patients with stroke; only 1 patient reached an exercise intensity over 30% heart rate reserve.²² Two other studies examined Lokomat walking with various degrees of robotic guidance, BWS, and walking speed, but neither reached considerably higher oxygen rates (4.89 to 8.09 mL/kg/min).^23,24 Likewise, walking in the Gait Trainer, an end-effector device, with 50% weight reduction induced similar rates of oxygen consumption (4.78-8.0 mL/kg/min). Without weight reduction, patients consumed slightly more energy (5.57-10.7 mL/kg/min).^25,26

Until now, the number of studies investigating oxygen consumption and cardiorespiratory demand during robot-assisted gait have been limited. Moreover, it is not yet known whether the physiological responses and perceived exertion change with respect to the duration of walking. Although, in the literature, gait training often lasts 30 minutes,^27,28 previous work only analyzed the physiological responses during short duration walking (3 to 6 minutes^22-26,29). However, decreased cardiorespiratory fitness levels after stroke² may result in a state in which the aerobic metabolism alone is unable to generate enough adenosine triphosphate (ATP)—the cells’ primary energy source—and the less-efficient anaerobic metabolism is needed to supplement ATP formation.³⁰ If the exercise intensity of gait training is around the patients’ anaerobic threshold, the anaerobic metabolism will raise lactate levels, which will elevate carbon dioxide production as a result of lactic acid buffering, and trigger the respiratory system to increase the amount of air exhaled per minute (ie, minute ventilation).³¹ The increased lactate production and carbohydrate consumption may result in an increased susceptibility to fatigue.

The aims of this study were 3-fold: (1) to compare the physiological responses and perceived exertion of robot-assisted gait with body weight–supported treadmill and overground walking, (2) to investigate if the physiological responses and perceived exertion change with respect to the duration of walking, and (3) to investigate if the physiological responses and perceived exertion reach aerobic training recommendations. This study is the first to examine poststroke physiological responses and perceived exertion during robot-assisted and body weight–supported gait for 30 minutes.

Methods

Study Design

This study was an experimental, cross-sectional study with randomized crossover design. The study protocol was registered prospectively (NCT02680496) and approved by the medical ethics committees of the University Hospital Brussels and Jessa Hospital Hasselt (B U N 143201526040).

Participants

We recruited 14 participants in the Rehabilitation Centre St Ursula (Herk-de-Stad, Belgium) from February 2016 to July 2017. All participants gave written consent to participate in this trial.

We included adult stroke survivors up to 1 year poststroke, who had a lower-limb motor impairment assessed by visual inspection (eg, poor foot clearance, knee instability, excessive hip circumduction, or gait asymmetry). To analyze training durations up to 30 minutes, we selected participants who could walk overground (functional ambulation category [FAC] 2-4) for at least 10 minutes at self-selected comfortable walking speed (CWS); BWS was allowed (LiteGait, Mobility Research, Tempe, AZ).

We excluded stroke patients with contraindications to exercise testing according to the American College of Sports Medicine,³² contraindications for Lokomat training,³³ musculoskeletal problems—other than stroke—affecting the ability to walk, concurrent pulmonary diseases, concurrent neurological diseases, communicative and/or cognitive problems affecting the ability to follow or comprehend instructions, and other problems affecting the execution of the interventions (eg, severe spasticity or dermatological conditions).

Experimental Walking Conditions

Participants completed 3 randomized walking trials within 1 week (Monday, Wednesday, and Friday between 3 and 5 pm): walking in the Lokomat with 60% guidance force (percentage frequently used in the initial phase of therapy²³), walking on a treadmill, and walking overground. For each participant, we determined CWS and the required level of BWS during an entry screening test (10-minute walking test on a 30-m indoor track with the use of the LiteGait BWS system). This speed and percentage of BWS were used in each condition. CWS was set at a maximum of 3.2 km/h (ie, maximum Lokomat speed).³³ Each condition lasted 30 minutes (if possible). Patients were instructed to hold the handrails. Walking tests were terminated early when participants were unable to continue walking or presented relative or absolute indications.³⁴

Physiological Measures and Rating of Perceived Exertion

At the start of each walking condition, a flexible facemask (small/medium, Cortex, Germany), lightweight chest-carrying gas analysis system (MetaMax 3B, Cortex, Germany), and heart rate belt (Polar H7, Polar Electro, Finland) were fitted. Participants were seated for 5 minutes of rest and afterward installed in one of the walking conditions (Figure 1). Breath-by-breath gas analysis and heart rate monitoring were conducted continuously from the beginning of rest to the end of walking. At the end of rest and at the end of every minute during walking, we asked participants to point their rating of perceived exertion on a 6 to 20 Borg Scale (6 representing no exertion; 20 representing maximal exertion).

Figure 1.

Experimental setup of the walking conditions: A. Lokomat V5.0 with Levi body weight support system. B. Woodway treadmill with Levi body weight support system. C. 30-m Overground track with LiteGait body weight support system.

Before the start of each walking condition, gas and volume calibrations were performed in accordance with the manufacturer’s instructions. We asked participants not to consume food, alcohol, caffeine, or nicotine at least 3 hours prior to the interventions and not to perform additional strenuous activities (apart from their regular therapy) at least 12 hours prior to the interventions.³² Usual medication intake was allowed with small amounts of water.^32,34

Outcomes

To investigate the effect on physiological responses, we analyzed: net oxygen consumption (mL/kg/min), net carbon dioxide production (mL/kg/min), net heart rate (beats/min), net minute ventilation (L/min), net breathing frequency (breaths/min), net respiratory exchange ratio (no unit), and net rating of perceived exertion (0-14). We analyzed net values, because net values take into account testing conditions (eg, variation in environmental factors or personal routines) and show less variability.^35,36 Net values were obtained by subtracting resting values (median of minute 5) from gross values. The first 5 minutes of walking were used to reach a steady-state condition and were not used in the analyses. Minutes 6, 18, and 30 were statistically analyzed (see below).

To compare the exercise intensity with aerobic guidelines for stroke survivors (ie, 40%-70% heart rate reserve, 55%-80% maximum heart rate, 11-14 rating of perceived exertion),⁹ we calculated: percentage of predicted heart rate reserve {[(Heart rate_walking − Heart rate_rest)/(Maximum heart rate_predicted − Heart rate_rest) × 100]}, percentage of predicted maximum heart rate [(Heart rate_walking/Maximum heart rate_predicted) × 100], and gross rating of perceived exertion. Because we could not perform incremental exercise tests to determine the participants’ maximum heart rate, we predicted maximum heart rate using the following formulas: 164 − 0.7 × Age (for persons taking β-blocking medication³⁷) or 206.9 − 0.67 × Age (for persons not taking β-blocking medication³²). Because predictions for people after stroke are not available, we adopted formulas for people with heart failure and healthy adults.^38,39

Data Processing

We found no universal standard for the data processing of breath-by-breath data. A common procedure in the corresponding literature is time averaging over 30- to 120-s intervals.^23-26,29 We chose to calculate medians over 60-s intervals because medians are less affected by irrelevant outliers caused by coughing or talking unintentionally during perceived exertion scoring.

Statistical Analysis

A 3 × 3 2-way repeated-measures ANOVA was conducted to compare the main effects of condition (Lokomat, treadmill, overground) and duration (6, 18, 30 minutes) with the interaction effect between condition and duration on net oxygen consumption, net carbon dioxide production, net heart rate, net minute ventilation, net breathing frequency, net respiratory exchange ratio, and net rating of perceived exertion (SPSS, IBM, Chicago, IL). Significance level was set at 5%. Bonferroni corrections were applied in post hoc analyses to adjust for multiple pairwise comparisons. Effect sizes were calculated to interpret the magnitude of effects: omega-squared (ω²) for repeated-measures ANOVA⁴⁰ and Hedges’ g_average for pairwise comparisons.⁴¹ Effect sizes were rated small (ω² = 0.01 and Hedges’ g_average = 0.2), medium (ω² = 0.06 and Hedges’ g_average = 0.5), or large (ω² = 0.14 and Hedges’ g_average = 0.8). The Shapiro-Wilk test and Q-Q plots were interpreted to check normality and the Mauchly’s test to test sphericity. If sphericity was violated, the Greenhouse-Geisser (ε < 0.75) or Huynh-Feldt (ε > 0.75) correction was applied.⁴⁰ One-sample t-tests were used to test whether percentage of heart rate reserve, percentage of predicted maximum heart rate, and gross rating of perceived exertion were above minimum and below maximum threshold for aerobic exercise at minutes 6, 18, and 30. Bonferroni corrections were applied to adjust for multiple testing.

Results

Participant Flow and Characteristics

A total of 14 participants were enrolled in this study; 8 participants completed all three 30-minute walking trials. Table 1 lists the demographic and clinical characteristics of each participant, including reasons for dropping out.

Table 1.

Demographic and Clinical Characteristics of Included Participants With Reasons for Dropping Out.

Sex	Age (years)	Weight (kg)	Height (m)	BMI (kg/m²)	Time Since Stroke (Months)	Type of Stroke	Hemiplegic Side	Walking Aid	β-Blockers	FAC (0-5)	CWS (km/h)	BWS (%)	Lokomat Experience (Sessions)
M	62	76	1.70	26	3.2	Ischemic	Left	Walking stick	No	4	2.0	50	0
M	48	86	1.80	27	10.8	Hemorrhagic	Left	Walking stick	No	4	2.0	40	32
M	63	80	1.80	25	1.4	Ischemic	Right	Walking stick	No	4	1.8	30	0
F	59	107	1.60	42	0.9	Hemorrhagic	Left	Walker	No	3	2.0	40	0
M	74	74	1.60	29	1.1	Ischemic	Right	Walker	Yes	3	1.7	40	0
M	53	86	1.68	30	2.0	Hemorrhagic	Left	Rollator	Yes	3	1.7	20	0
M	58	65	1.71	22	1.0	Ischemic	Right	No	Yes	4	2.2	50	0
M	59	104	1.73	35	5.9	Ischemic	Left	Walking stick	Yes	4	1.5	40	17
M^a	48	74	1.75	24	4.6	Ischemic	Right	No	Yes	4	1.9	40	5
M^b	64	104	1.73	35	1.0	Ischemic	Left	No	No	4	3.1	0	0
M^c	73	77	1.71	26	0.7	Ischemic	Right	Rollator	No	3	1.5	60	1
M^d	59	86	1.77	27	0.7	Ischemic	Right	No	No	4	2.6	50	0
F^e	74	58	1.60	23	2.9	Ischemic	Left	Wheelchair	Yes	3	2.0	50	3
M^f	55	70	1.80	22	3.4	Hemorrhagic	Right	Wheelchair	Yes	2	2.0	60	6

Abbreviations: BMI, body mass index; BWS, body weight support; CWS, comfortable walking speed; F, female; FAC, functional ambulation category; M, male.

Dropped out after 23 minutes of treadmill walking because of fatigue (data not included in statistical analyses).

Dropped out after 2 minutes Lokomat walking because of severe pain at the hemiplegic leg (data not included in statistical analyses).

Continuous blocking of the Lokomat system prevented walking (data not included in statistical analyses).

Dropped out after 8 minutes treadmill walking because of the inability to maintain the walking speed and after 19 minutes overground walking because of fatigue (data not included in statistical analyses).

Inability to reach walking speed prevented treadmill walking; dropped out after 17 minutes overground walking because of fatigue (data not included in statistical analyses).

Participants (12 male and 2 female) were between 48 and 74 years old (median: 59 years) and 1 to 11 months after stroke (median: 1.7 months). Eight participants walked independently on level surfaces (FAC 4); 5 required verbal supervision or stand-by help (FAC 3); and 1 needed support of a person for balance and coordination (FAC 2). Participants walked at a median CWS of 2.0 km/h in combination with a median BWS of 40%. Six participants previously walked in the Lokomat (1 to 32 sessions).

Effect of Walking Condition and Duration on Physiological Responses and Perceived Exertion

Because of listwise deletion, we could only include data of 8 participants. Table 2 describes the descriptive statistics for each outcome.

Table 2.

Physiological Responses and Perceived Exertion During Lokomat, Treadmill, and Overground Walking.^a

Outcome	Duration (minutes)	Condition
		Lokomat				Treadmill				Overground
		n	M	SD	95% CI	n	M	SD	95% CI	n	M	SD	95% CI
Gross VO₂ (mL/kg/min)	6	8	5.91	0.68	[5.34, 6.47]	8	8.89	1.51	[7.63, 10.16]	8	8.87	1.91	[7.27, 10.47]
	18 min	8	6.00	1.13	[5.06, 6.95]	8	8.72	1.55	[7.42, 10.01]	8	8.87	1.75	[7.41, 10.34]
	30	8	6.45	0.91	[5.69, 7.21]	8	8.92	1.60	[7.58, 10.26]	8	8.81	1.90	[7.21, 10.40]
Net VO₂ (mL/kg/min)	6	8	2.54	0.75	[1.91, 3.17]	8	5.51	1.32	[4.41, 6.62]	8	5.47	1.61	[4.13, 6.82]
	18	8	2.63	1.71	[1.20, 4.06]	8	5.34	1.30	[4.26, 6.42]	8	5.47	1.36	[4.33, 6.61]
	30	8	3.08	1.34	[1.96, 4.20]	8	5.54	1.38	[4.39, 6.69]	8	5.41	1.63	[4.04, 6.77]
Gross VCO₂ (mL/kg/min)	6	8	5.02	0.70	[4.44, 5.60]	8	7.94	1.44	[6.49, 8.90]	8	7.53	1.76	[6.06, 9.00]
	18	8	5.13	1.02	[4.27, 5.98]	8	7.63	1.45	[6.41, 8.84]	8	7.84	1.52	[6.57, 9.11]
	30	8	5.37	0.83	[4.68, 6.07]	8	7.88	1.44	[6.68, 9.08]	8	7.76	1.68	[6.36, 9.17]
Net VCO₂ (mL/kg/min)	6	8	2.08	0.79	[1.42, 2.74]	8	4.84	1.35	[3.71, 5.97]	8	4.53	1.38	[3.38, 5.69]
	18	8	2.19	1.62	[0.83, 3.55]	8	4.77	1.23	[3.74, 5.81]	8	4.84	1.12	[3.91, 5.76]
	30	8	2.44	1.21	[1.43, 3.45]	8	5.02	1.24	[3.98, 6.06]	8	4.77	1.41	[3.59, 5.95]
Gross HR (beats/min)	6	8	77	11	[68, 86]	8	86	12	[76, 96]	8	87	11	[78, 96]
	18	8	76	7	[70, 82]	8	87	13	[76, 98]	8	88	12	[78, 98]
	30	8	78	8	[72, 85]	8	91	16	[77, 104]	8	88	12	[78, 98]
Net HR (beats/min)	6	8	12	8	[5, 18]	8	18	6	[13, 22]	8	19	10	[10, 27]
	18	8	10	7	[4, 16]	8	19	6	[13, 24]	8	20	9	[13, 28]
	30	8	13	6	[8, 18]	8	23	9	[15, 30]	8	20	7	[14, 26]
Gross VE (L/min)	6	8	16.63	1.75	[15.17, 18.09]	8	22.87	4.37	[19.21, 26.52]	8	22.79	4.53	[19.00, 26.57]
	18	8	17.31	3.71	[14.21, 20.41]	8	23.00	3.88	[19.76, 26.25]	8	24.14	4.95	[20.00, 28.28]
	30	8	18.57	3.09	[15.98, 21.15]	8	24.80	4.25	[21.25, 28.35]	8	24.38	5.43	[19.84, 28.92]
Net VE (L/min)	6	8	6.45	1.67	[5.06, 7.06]	8	13.17	4.12	[9.73, 16.62]	8	12.34	3.72	[9.23, 15.45]
	18	8	7.14	4.80	[3.12, 11.15]	8	13.31	3.65	[10.26, 16.36]	8	13.69	4.00	[10.34, 17.04]
	30	8	8.39	3.92	[5.11, 11.66]	8	15.11	4.13	[11.65, 18.56]	8	13.93	4.78	[9.93, 17.92]
Gross BF (breaths/min)	6	8	22	3	[19, 25]	8	23	3	[20, 25]	8	24	6	[19, 29]
	18	8	23	4	[20, 26]	8	24	3	[21, 27]	8	25	7	[19, 31]
	30	8	24	3	[21, 27]	8	23	4	[20, 27]	8	27	8	[20, 34]
Net BF (breaths/min)	6	8	6	3	[4, 9]	8	6	2	[5, 8]	8	7	4	[4, 11]
	18	8	7	4	[4, 10]	8	8	3	[5, 10]	8	9	5	[4, 13]
	30	8	8	3	[5, 11]	8	7	3	[5, 10]	8	10	7	[4, 15]
Gross RER (no unit)	6	8	0.85	0.05	[0.81, 0.89]	8	0.86	0.04	[0.83, 0.90]	8	0.85	0.04	[0.81, 0.88]
	18	8	0.85	0.03	[0.83, 0.88]	8	0.87	0.02	[0.86, 0.89]	8	0.88	0.03	[0.86, 0.91]
	30	8	0.83	0.03	[0.81, 0.85]	8	0.88	0.03	[0.86, 0.91]	8	0.88	0.03	[0.85, 0.91]
Net RER (no unit)	6	8	−0.02	0.06	[−0.07, 0.04]	8	0.02	0.04	[−0.02, 0.05]	8	−0.03	0.03	[−0.06, 0.00]
	18	8	−0.01	0.06	[−0.06, 0.04]	8	0.03	0.06	[−0.02, 0.07]	8	0.01	0.04	[−0.03, 0.04]
	30	8	−0.03	0.04	[−0.06, 0.00]	8	0.04	0.06	[−0.01, 0.09]	8	0.003	0.03	[−0.02, 0.03]
Gross RPE (6-20)	6	11	7	3	[6, 9]	11	8	2	[7, 9]	12	8	2	[6, 9]
	18	11	9	4	[6, 11]	11	10	2	[8, 11]	12	10	3	[8, 12]
	30	11	10	4	[8, 13]	11	11	3	[9, 14]	12	11	4	[9, 14]
Net RPE (0-14)	6	8	1	2	[−0.6, 2.8]	8	2	2	[0.5, 3.5]	8	2	2	[0.0, 3.2]
	18	8	3	3	[0.2, 4.8]	8	4	2	[2.3, 5.7]	8	4	3	[1.5, 6.8]
	30	8	4	4	[0.8, 7.0]	8	6	3	[3.6, 8.4]	8	6	3	[3.4, 9.1]
Percentage HRR (%)	6	11	16.8	14.3	[7.2, 26.4]	11	23.9	7.0	[19.2, 28.6]	12	33.1	20.7	[19.9, 46.3]
	18	11	14.9	10.9	[7.5, 22.2]	11	25.8	7.3	[20.8, 30.7]	12	35.7	21.7	[21.9, 49.5]
	30	11	17.7	10.6	[10.6, 24.9]	11	30.2	12.0	[22.1, 38.3]	12	36.7	20.9	[23.4, 50.0]
Percentage HR_max (%)	6	11	54.1	12.7	[45.5, 62.6]	11	60.2	9.4	[53.9, 66.5]	12	64.4	14.5	[55.2, 63.7]
	18	11	53.2	11.1	[45.7, 60.7]	11	61.2	9.0	[55.2, 67.3]	12	65.6	14.6	[56.3, 74.9]
	30	11	54.7	11.0	[47.4, 62.1]	11	63.4	10.2	[56.5, 70.3]	12	66.2	14.1	[57.2, 75.2]

Abbreviations: BF, breathing frequency; HR, heart rate; HRmax, predicted maximum heart rate; HRR, heart rate reserve; M, mean; RER, respiratory exchange ratio; RPE, rating of perceived exertion; VCO₂, carbon dioxide production; VE, minute ventilation; VO₂, oxygen consumption.

Net values were obtained by subtracting resting values (median of minute 5) from gross values.

We found no significant interaction effects between condition and duration, except for net respiratory exchange ratio; the magnitude of the effect was small (Supplementary Material 1, available online). Post hoc tests indicated that net respiratory exchange ratio significantly increased between 6 and 18 minutes and between 6 and 30 minutes overground walking; the magnitudes of the effects were large (Supplementary Material 2, available online). No significant changes were found during Lokomat and treadmill walking.

The main effect of condition was significant for net oxygen consumption, net carbon dioxide production, net heart rate, and net minute ventilation; the magnitudes of the effects were large (Supplementary Material 1, available online). Post hoc tests showed that net oxygen consumption, net carbon dioxide production, net heart rate, and net minute ventilation were significantly lower during Lokomat walking compared with treadmill and overground walking. The magnitudes of the effects were large. No significant differences were found between treadmill and overground walking (Supplementary Material 3, available online).

The main effect of duration was significant for net heart rate, net minute ventilation, net breathing frequency, and net rating of perceived exertion; the magnitudes of the effects were small to large (Supplementary Material 1, available online). Post hoc tests showed that net heart rate significantly increased between 18 and 30 minutes; net minute ventilation significantly increased between 6 and 30 minutes, and between 18 and 30 minutes; and net breathing frequency significantly increased between 6 and 18 minutes, and between 6 and 30 minutes. The magnitudes of the effects were small. Net rating of perceived exertion significantly increased between each duration; the magnitude of the effect was large (Supplementary Material 4, available online). Figure 2 illustrates these results.

Figure 2.

Net physiological responses and perceived exertion during 30-minute Lokomat, treadmill, and overground training. Means and 95% CIs are plotted (n = 8).

Comparison With Aerobic Training Recommendations

Lokomat Walking

Throughout Lokomat walking, mean percentage of heart rate reserve was significantly below the minimum threshold for aerobic exercise, whereas mean percentage of predicted maximum heart rate was at the lower border. Mean gross rating of perceived exertion increased toward the aerobic target zone by the end (Figure 3; Supplementary Material 5 available online).

Figure 3.

Exercise intensity of Lokomat, treadmill, and overground training compared with aerobic training recommendations for stroke survivors. Means and 95% CIs are plotted as well as aerobic training recommendations (gray zones).

Treadmill Walking

Throughout treadmill walking, mean percentage of predicted maximum heart rate fell within the aerobic target zone, and mean percentage of heart rate reserve was below the minimum threshold. Halfway to the end, mean gross rating of perceived exertion increased toward the aerobic target zone (Figure 3; Supplementary material 5, available online).

Overground Walking

From the beginning to the end, mean percentage of predicted maximum heart rate fell within the aerobic target zone, whereas mean percentage of predicted heart rate reserve was at the lower border. Perceived exertion increased during overground walking and reached aerobic guidelines halfway to the end (Figure 3; Supplementary Material 5, available online).

Discussion

This study compared physiological responses and perceived exertion between Lokomat, treadmill, and overground walking at matched speeds and levels of BWS and investigated the effect of prolonged walking duration on physiological responses and perceived exertion.

In our sample, net oxygen consumption, net carbon dioxide production, net heart rate, and net minute ventilation were about half as high during 30-minute robot-assisted gait than during 30-minute body weight–supported treadmill and overground walking. These results were in line with van Nunen et al,²² who reported that Lokomat walking was less intensive than 6-minute unsupported overground walking with usual walking aids (patients consumed 27% less oxygen), even though patients walked much slower overground.²² Lower physiological responses during Lokomat walking may be explained by a general reduction in muscle activity of the lower-limb muscles (both at the affected and nonaffected side), resulting in an overall reduction in active contribution.⁴² Stroke survivors also consume less energy when they walk more symmetrically.⁴³ Lokomat walking reduces temporal asymmetry, which occurs during treadmill walking, because of a prolongation of the relative single support duration in the affected limb.⁴²

By the end, patients took slightly deeper breaths (MD = 1.58 L/min; P < .05) and breathed a bit faster (MD = 2 breaths/min; P < .05), somewhat demonstrating that the anaerobic metabolism supplements the ATP formation and triggers the respiratory system to remove excess carbon dioxide. However, the differences were small—clinically irrelevant—and we did not detect a significant rise in the net carbon dioxide production. The net respiratory exchange ratio slightly increased during prolonged overground walking (MD = 0.03; P < .05), somewhat indicating that less fats and more carbohydrates are being metabolized to supply the body with energy. The reason why we did not detect a rise is because the exercise intensity of body weight–supported walking (~8.9 mL/kg/min) and especially robot-assisted walking (~6.0 mL/kg/min) may have been below these patients’ anaerobic threshold. In the literature, the ventilatory threshold of patients with stroke was around 9.6 ± 0.4 mL/kg/min in subacute patients (n = 34)⁴⁴ and around 10.6 ± 2.8 mL/kg/min in chronic patients (n = 59).⁴⁵

Although our patients consumed less energy, they felt equally as tired during robot-assisted gait as during body weight–supported gait. Nevertheless, patients felt more tired when the walking duration increased. Patients perceived that walking was two and a half times more strenuous by the end. To our knowledge, only a single study investigated whether physical exertion increased with respect to the duration of overground walking after stroke (6- vs 18-minute walk).⁴⁶ Carver et al⁴⁶ demonstrated that patients with stroke were tired at the end of an 18-minute walk (shifted from very light to somewhat hard perceived exertion), but they were able to sustain their effort throughout the walk (ie, they kept their walking speed constant).⁴⁶ In our sample, we found an analogous—yet smaller—increase in perceived exertion (shift from extremely light to light perceived exertion).

Overall, the exercise intensity of Lokomat walking was below current exercise recommendations for stroke survivors—only mean percentage of maximum heart rate was at the lower bound of the minimum threshold. The exercise intensity of body weight–supported treadmill and overground walking was at the lower border or within the aerobic training zone, except for percentage of heart rate reserve during treadmill walking. This result was in line with that of van Nunen et al²² who found a similar mean percentage of predicted heart rate reserve during Lokomat walking (22% ± 7%).²² Although these results suggest that according to current stroke guidelines the exercise intensity of Lokomat walking is insufficient to induce an aerobic training effect, Chang et al⁴⁷ found that 2-week Lokomat training improved peak oxygen consumption by 13% in patients with subacute stroke. This discrepancy may imply that in patients with stroke, light-intensity exercise may be sufficient to induce a cardiorespiratory training effect (requiring a revision of the current exercise guidelines for stroke survivors) or that the exercise intensity of Lokomat training is currently underestimated.

Oxygen consumption and heart rate levels in our patients during Lokomat walking were within the ranges of those observed in the literature.^22-24 We found somewhat lower oxygen rates than those observed by both Krewer et al²³ and van Nunen et al,²² but in those studies, patients walked with a lower percentage of BWS (~30%). Because walking with lower loads requires less oxygen consumption, we can probably designate dissimilarities to the differences in loading.^23,26 In the study of van Nunen et al,²² patients also walked with a lower percentage of robotic guidance (~30%).²² However, this lower percentage would probably not explain the higher oxygen rates, because lowering the robotic guidance in previous studies either left oxygen uptake unchanged or induced a decline.^23,24 The walking ability of these patients, however, was a bit lower than the ability of our patients (mean FAC 2.6 vs 3.5). This may also add to the higher oxygen rates.⁴⁸ We found slightly higher oxygen rates and heart rates than those observed by Lee et al,²⁴ but in their study, patients walked with a fixed percentage of BWS (50%). In our study, the percentage was individually adjusted (as low as possible, with an average of 41%). Using an individual approach could make walking more intensive, whereas using a fixed percentage could provide too much support for some patients and too little for others. Another reason why we found slightly higher rates may be a result of the degree of familiarization with the Lokomat. In our study, not all patients were familiarized with the Lokomat, whereas in the study of Lee et al,²⁴ patients trained 3 times a week for 2 weeks prior to the experiments.²⁴ Sawicki and Ferris⁴⁹ demonstrated in healthy individuals that the metabolic benefit of waking with a powered ankle exoskeleton increased with practice: they found a 22% reduction in metabolic power between the beginning of the first and the end of the third session.

Future studies should further explore how robotic devices can be exploited to induce aerobic training after stroke. Currently, the highest oxygen demands have been achieved by combining higher walking speeds (1.8 km/h or higher) with lower levels of BWS (50% or lower).^23,24 Nonetheless, the effect of changing the settings appears to be quite small. Stoller et al⁵⁰ were able to reach 26% higher exercise intensities (in terms of percentage of heart rate reserve) by using progressive feedback-controlled robot-assisted treadmill training rather than conventional robot-assisted treadmill training providing constant physical support by the actuated exoskeleton. The feedback-controlled approach was based on the forces applied on the exoskeleton by the participant. Participants had to keep the visualized work rate as close as possible to the target work rate by volitional muscle activity.⁵¹ Although this study showed promising results—they found a significant overall increase in cardiorespiratory fitness after a 4-week training period early after stroke—the majority of participants (4 out of 7) in the feedback-controlled group still did not reach the aerobic threshold (ie, 40% heart rate reserve).⁵⁰ Future research should further examine the effect of variable workload and asymmetrical settings as well as overground exoskeleton walking. Current evidence is based on static robotic devices in combination with BWS systems (ie, the Lokomat and the Gait Trainer),^22-26 but it is possible that the exercise intensity of walking with wearable exoskeletons—dynamic overground systems that do not provide weight reduction—may induce higher cardiometabolic demands.

This study has a couple of limitations. A first limitation is our limited sample size, which compromises the generalizability of our results—a problem that often occurs in this area of research.²⁰ This is not unusual because recruiting patients for studies like these is challenging. A previous study reported that it is hard to convince patients, who already face many problems and discomforts, to engage in studies requiring active participation, food restriction, breathing through a facemask, and so on.⁵² Nevertheless, in future research, we should keep investing in obtaining large samples to be able to draw well-founded conclusions.

A second limitation is the possible under- or overestimation of maximum heart rate, affecting comparisons with aerobic training recommendations. Because we were unable to perform incremental exercise tests and predictions for people after stroke are not yet available, we predicted participants’ maximum heart rate based on formulas for people with heart failure and people taking β-blocking medication. Although these predictions are quite often used,^22,38,39 they may be less accurate for people with stroke. Future studies should invest in validating predictions for stroke survivors. Furthermore, physiological resting values were measured pre-exercise, possibly causing excitement or stress (similar to a white-coat effect⁵³), slightly overestimating true resting values and underestimating percentages of heart rate reserve.

A third limitation is a possible discrepancy between CWS overground, on a treadmill, and in a Lokomat-system. In this study, we defined patients’ CWS during an overground walking test. This speed was used in the different walking conditions, although stroke survivors often walk faster overground than on a treadmill.⁵⁴ This discrepancy may be one reason why the Lokomat system kept blocking in one patient and data could not be used. This patient reached a high overground CWS (3.2 km/h),⁵⁴ matching the maximum Lokomat speed. During Lokomat walking, the pelvis is secured in a harness and fixed between 2 hip cushions, leading to restrictions in the amplitudes of thorax and pelvis movements.^55,56 Higher walking speeds (starting between 2.7 and 3.6 km/h) have been associated with larger transverse pelvic and trunk rotations^57,58—movements that are hindered in the Lokomat.

Conclusion

Our results suggest that in ambulatory stroke survivors, the oxygen and cardiorespiratory demand during robot-assisted gait with 60% robotic guidance at constant work load is considerably lower than during overground or treadmill walking at matched speeds and levels of BWS (below aerobic thresholds). To train the cardiorespiratory system, it is important that patients who can walk (with or without BWS) train without the help of robotic devices (or that settings should be done differently). Nevertheless, lower physical requirements during robot-assisted gait can enable certain patients (in our case two) to train longer. Future studies should examine how robotic devices can be exploited to meet the exercise guidelines for stroke survivors.

Footnotes

Acknowledgements

The authors would like to thank the therapists of the Rehabilitation Centre St Ursula (Herk-de-Stad, Belgium) for managing the Lokomat system and recruiting participants; Ken Janssens, Nicolas Haegeman, Laura Vander Goten, Antoine Verbeylen, Nikkie Dassen, and Sam De Buyzer for their help with data gathering; and Patricia Van de Walle and Bas De Geus for their advice on breath-by-breath gas analyses.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: N.L. is a Strategic Basic Research fellow funded by the Research Foundation – Flanders (FWO).

Supplementary material for this article is available on the Neurore-habilitation & Neural Repair website along with the online version of the article.

ORCID iD

Nina Lefeber, .

References

Kelly

Kilbreath

Davis

Zeman

Raymond

Cardiorespiratory fitness and walking ability in subacute stroke patients. Arch Phys Med Rehabil. 2003;84:1780-1785.

Smith

Saunders

Mead

Cardiorespiratory fitness after stroke: a systematic review. Int J Stroke. 2012;7:499-510.

Billinger

Coughenour

Mackay-Lyons

Ivey

FM.

Reduced cardiorespiratory fitness after stroke: biological consequences and exercise-induced adaptations. Stroke Res Treat. 2012;2012:959120.

Myers

McAuley

Lavie

Despres

Arena

Kokkinos

Physical activity and cardiorespiratory fitness as major markers of cardiovascular risk: their independent and interwoven importance to health status. Prog Cardiovasc Dis. 2015;57:306-314.

Brønnum-Hansen

Davidsen

Thorvaldsen

; Danish MONICA Study Group. Long-term survival and causes of death after stroke. Stroke. 2001;32:2131-2136.

Pang

Charlesworth

Lau

Chung

RC.

Using aerobic exercise to improve health outcomes and quality of life in stroke: evidence-based exercise prescription recommendations. Cerebrovasc Dis. 2013;35:7-22.

Marsden

Dunn

Callister

Levi

Spratt

NJ.

Characteristics of exercise training interventions to improve cardiorespiratory fitness after stroke: a systematic review with meta-analysis. Neurorehabil Neural Repair. 2013;27:775-788.

Saunders

Sanderson

Hayes

et al . Physical fitness training for stroke patients. Cochrane Database Syst Rev. 2016;(3):CD003316.

Billinger

Arena

Bernhardt

et al . Physical activity and exercise recommendations for stroke survivors: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2014;45:2532-2553.

10.

Latham

Jette

Slavin

et al . Physical therapy during stroke rehabilitation for people with different walking abilities. Arch Phys Med Rehabil. 2005;86(12, suppl 2):S41-S50.

11.

Foley

McClure

Meyer

Salter

Bureau

Teasell

Inpatient rehabilitation following stroke: amount of therapy received and associations with functional recovery. Disabil Rehabil. 2012;34:2132-2138.

12.

Otterman

van der Wees

Bernhardt

Kwakkel

Physical therapists’ guideline adherence on early mobilization and intensity of practice at Dutch acute stroke units: a country-wide survey. Stroke. 2012;43:2395-2401.

13.

De Wit

Putman

Lincoln

et al . Stroke rehabilitation in Europe: what do physiotherapists and occupational therapists actually do? Stroke. 2006;37:1483-1489.

14.

National Stroke Foundation. Clinical guidelines for stroke management 2010. https://www.pedro.org.au/wp-content/uploads/CPG_stroke.pdf. Accessed October 20, 2018.

15.

Lindsay

Bayley

Hellings

Hill

Woodbury

Phillips

Canadian best practice recommendations for stroke care (updated 2008). CMAJ. 2008;179(12, suppl):S1-S25.

16.

MacKay-Lyons

Makrides

Cardiovascular stress during a contemporary stroke rehabilitation program: is the intensity adequate to induce a training effect?

Arch Phys Med Rehabil. 2002;83:1378-1383.

17.

Kuys

Brauer

Ada

Routine physiotherapy does not induce a cardiorespiratory training effect post-stroke, regardless of walking ability. Physiother Res Int. 2006;11:219-227.

18.

Prout

Mansfield

McIlroy

Brooks

Physiotherapists’ perspectives on aerobic exercise early after stroke: a preliminary study. Physiother Theory Pract. 2016;32:452-460.

19.

Morone

Paolucci

Cherubini

et al . Robot-assisted gait training for stroke patients: current state of the art and perspectives of robotics. Neuropsychiatr Dis Treat. 2017;13:1303-1311.

20.

Lefeber

Swinnen

Kerckhofs

The immediate effects of robot-assistance on energy consumption and cardiorespiratory load during walking compared to walking without robot-assistance: a systematic review. Disabil Rehabil Assist Technol. 2017;12:657-671.

21.

Barroso

Santos

Pons

Moreno

. Muscular activation and kinetic effects of robotic guidance force on human walking. In: Pons

Torricelli

Pajaro

, eds. Converging Clinical and Engineering Research on Neurorehabilitation. Vol 1. Berlin, Germany: Springer; 2013:787-791.

22.

van Nunen

Gerrits

de Haan

Janssen

TW.

Exercise intensity of robot-assisted walking versus overground walking in nonambulatory stroke patients. J Rehabil Res Dev. 2012;49:1537-1546.

23.

Krewer

Müller

Husemann

Heller

Quintern

Koenig

The influence of different Lokomat walking conditions on the energy expenditure of hemiparetic patients and healthy subjects. Gait Posture. 2007;26:372-377.

24.

Lee

Han

Kim

Chun

Lee

YK.

Can lowering the guidance force of robot-assisted gait training induce a sufficient metabolic demand in subacute dependent ambulatory patients with stroke?

Arch Phys Med Rehabil. 2017;98:695-700.

25.

David

Regnaux

Lejaille

Louis

Bussel

Lofaso

Oxygen consumption during machine-assisted and unassisted walking: a pilot study in hemiplegic and healthy humans. Arch Phys Med Rehabil. 2006;87:482-489.

26.

Delussu

Morone

Iosa

Bragoni

Traballesi

Paolucci

Physiological responses and energy cost of walking on the gait trainer with and without body weight support in subacute stroke patients. J Neuroeng Rehabil. 2014;11:54.

27.

Mehrholz

Thomas

Elsner

Treadmill training and body weight support for walking after stroke. Cochrane Database Syst Rev. 2017;(8):CD002840.

28.

Mehrholz

Thomas

Werner

Kugler

Pohl

Elsner

Electromechanical-assisted training for walking after stroke. Cochrane Database Syst Rev. 2017;(5):CD006185.

29.

Danielsson

Sunnerhagen

KS.

Oxygen consumption during treadmill walking with and without body weight support in patients with hemiparesis after stroke and in healthy subjects. Arch Phys Med Rehabil. 2000;81:953-957.

30.

Petterson

Kuchta

Snyder-Mackler

. Aerobic metabolism during exercise. In: Donatelli

, ed. Sports-Specific Rehabilitation. Saint Louis, MO: Churchill Livingstone; 2007:65-85.

31.

Plowman

Smith

. Anaerobic metabolism during xercise. In: Donatelli

, ed. Sports-Specific Rehabilitation. Saint Louis, MO: Churchill Livingstone; 2007:39-63.

32.

American College of Sports Medicine. ACSM’s Guidelines for Exercise Testing and Prescription. Philadelphia, PA: Lippincott Williams & Wilkins; 2013.

33.

Schmidt

Hesse

Bernhardt

Krüger

HapticWalker—a novel haptic foot device. ACM Trans Appl Percept. 2005;2:166-180.

34.

Fletcher

Ades

Kligfield

et al . Exercise standards for testing and training: a scientific statement from the American Heart Association. Circulation. 2013;128:873-934.

35.

Baker

Hausch

McDowell

Reducing the variability of oxygen consumption measurements. Gait Posture. 2001;13:202-209.

36.

Dawes

Collett

Ramsbottom

Howells

Sackley

Wade

Measuring oxygen cost during level walking in individuals with acquired brain injury in the clinical setting. J Sports Sci Med. 2004;3:76-82.

37.

Brawner

Ehrman

Schairer

Cao

Keteyian

SJ.

Predicting maximum heart rate among patients with coronary heart disease receiving beta-adrenergic blockade therapy. Am Heart J. 2004;148:910-914.

38.

Billinger

Boyne

Coughenour

Dunning

Mattlage

Does aerobic exercise and the FITT principle fit into stroke recovery?

Curr Neurol Neurosci Rep. 2015;15:519.

39.

Kelly

Devasahayam

Chaves

et al . Intensifying functional task practice to meet aerobic training guidelines in stroke survivors. Front Physiol. 2017;8:809.

40.

Field

AP.

Discovering Statistics Using SPSS. 3rd ed. Thousand Oaks, CA: Sage; 2009.

41.

Lakens

Calculating and reporting effect sizes to facilitate cumulative science: a practical primer for t-tests and ANOVAs. Front Psychol. 2013;4:863.

42.

van Kammen

Boonstra

van der Woude

LHV

Reinders-Messelink

den Otter

Differences in muscle activity and temporal step parameters between Lokomat guided walking and treadmill walking in post-stroke hemiparetic patients and healthy walkers. J Neuroeng Rehabil. 2017;14:32.

43.

Awad

Palmer

Pohlig

Binder-Macleod

Reisman

DS.

Walking speed and step length asymmetry modify the energy cost of walking after stroke. Neurorehabil Neural Repair. 2015;29:416-423.

44.

Tang

Sibley

Thomas

McIlroy

Brooks

Maximal exercise test results in subacute stroke. Arch Phys Med Rehabil. 2006;87:1100-1105.

45.

Boyne

Reisman

Brian

et al . Ventilatory threshold may be a more specific measure of aerobic capacity than peak oxygen consumption rate in persons with stroke. Top Stroke Rehabil. 2017;24:149-157.

46.

Carver

Nadeau

Leroux

Relation between physical exertion and postural stability in hemiparetic participants secondary to stroke. Gait Posture. 2011;33:615-619.

47.

Chang

Kim

Huh

Lee

Kim

YH.

Effects of robot-assisted gait training on cardiopulmonary fitness in subacute stroke patients: a randomized controlled study. Neurorehabil Neural Repair. 2012;26:318-324.

48.

Ijmker

Houdijk

Lamoth

et al . Effect of balance support on the energy cost of walking after stroke. Arch Phys Med Rehabil. 2013;94:2255-2261.

49.

Sawicki

Ferris

DP.

Mechanics and energetics of level walking with powered ankle exoskeletons. J Exp Biol. 2008;211(pt 9):1402-1413.

50.

Stoller

de Bruin

Schindelholz

Schuster-Amft

de Bie

Hunt

KJ.

Efficacy of feedback-controlled robotics-assisted treadmill exercise to improve cardiovascular fitness early after stroke: a randomized controlled pilot trial. J Neurol Phys Ther. 2015;39:156-165.

51.

Stoller

Schindelholz

Bichsel

et al . Feedback-controlled robotics-assisted treadmill exercise to assess and influence aerobic capacity early after stroke: a proof-of-concept study. Disabil Rehabil Assist Technol. 2014;9:271-278.

52.

Delussu

Morone

Iosa

Bragoni

Paolucci

Traballesi

Concurrent validity of Physiological Cost Index in walking over ground and during robotic training in subacute stroke patients. Biomed Res Int. 2014;2014:384896.

53.

Lequeux

Uzan

Rehman

MB.

Does resting heart rate measured by the physician reflect the patient’s true resting heart rate? White-coat heart rate. Indian Heart J. 2018;70:93-98.

54.

Bayat

Barbeau

Lamontagne

Speed and temporal-distance adaptations during treadmill and overground walking following stroke. Neurorehabil Neural Repair. 2005;19:115-124.

55.

Swinnen

Baeyens

Knaepen

et al . Robot-assisted walking with the Lokomat: the influence of different levels of guidance force on thorax and pelvis kinematics. Clin Biomech (Bristol, Avon). 2015;30:254-259.

56.

Swinnen

Baeyens

Knaepen

et al . Walking with robot assistance: the influence of body weight support on the trunk and pelvis kinematics. Disabil Rehabil Assist Technol. 2015;10:252-257.

57.

LaFiandra

Wagenaar

Holt

Obusek

JP.

How do load carriage and walking speed influence trunk coordination and stride parameters?

J Biomech. 2003;36:87-95.

58.

Wagenaar

Beek

WJ.

Hemiplegic gait: a kinematic analysis using walking speed as a basis. J Biomech. 1992;25:1007-1015.

Physiological Responses and Perceived Exertion During Robot-Assisted and Body Weight–Supported Gait After Stroke

Abstract

Keywords

Introduction

Methods

Study Design

Participants

Experimental Walking Conditions

Physiological Measures and Rating of Perceived Exertion

Outcomes

Data Processing

Statistical Analysis

Results

Participant Flow and Characteristics

Effect of Walking Condition and Duration on Physiological Responses and Perceived Exertion

Comparison With Aerobic Training Recommendations

Lokomat Walking

Treadmill Walking

Overground Walking

Discussion

Conclusion

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

ORCID iD

References