ORIGINAL REPORT

METABOLIC COST AND MECHANICAL EFFICIENCY OF A NOVEL HANDLE-BASED DEVICE FOR WHEELCHAIR PROPULSION

Markus PUCHINGER¹, MSc, Nithin KURUP¹, PhD, Karin GSTALTNER², Marcus G. PANDY³, PhD and Margit GFÖHLER¹, PhD

From the ¹Research Division of Biomechanics and Rehabilitation Engineering, Department of Engineering Design and Product Development, Technische Universität Wien (TU Wien), ²AUVA Rehabilitation Centre Meidling, Vienna, Austria and ³Department of Mechanical Engineering, University of Melbourne, Parkville, Australia

Objective: To investigate differences in metabolic cost and gross mechanical efficiency of a novel handle-based wheelchair propulsion device and to compare its performance with conventional push-rim propulsion.

Design: Double-group comparative study between 2 different propulsion methods.

Participants: Eight paraplegic individuals and 10 non-disabled persons.

Methods: Participants performed the same exercise using a push-rim device and the novel handle-based device on a wheelchair-based test rig. The exercise consisted of a combined submaximal and maximal test. Power output, oxygen uptake, ventilation, respiratory exchange ratio and heart rate were recorded continuously during the tests. Analysis of variance was performed to determine the effects of group, mode and on power output.

Results: Submaximal exercise resulted in a higher efficiency for the novel device and significant main effects of propulsion mode on all investigated parameters, except heart rate. On the respiratory exchange ratio, a significant interaction effect was found for both mode and group. The maximal exercise resulted in a higher peak power output and lower peak heart rate during propulsion using the handle-based device. A significant main effect on mode for mean peak power output, ventilation and heart rate was also observed.

Conclusion: Wheelchair propulsion using the handle-based device resulted in lower physical responses and higher mechanical efficiency, suggesting that this novel design may be well suited for indoor use, thereby offering an attractive alternative to push-rim wheelchairs.

LAY ABSTRACT

The push-rim is the preferred mode of propulsion for more than 90% of all self-propelled wheelchair users, even though it is the least efficient. Furthermore, push-rim propulsion is highly strenuous for the musculoskeletal system and often leads to severe upper limb injuries. Alternative modes of manual wheelchair propulsion are available (e.g. arm-crank propulsion (handbikes) and lever-propulsion) but most of these are bulky, heavy and mostly suitable for outdoor use. The aim of the current study was to investigate differences in metabolic cost and mechanical efficiency for a novel handle-based and ergonomically optimized device and to compare its performance with conventional push-rim propulsion. Eight paraplegic subjects and 10 non-disabled controls performed exercises at different power resistances. The results show that the performance of the handle-based device is below that of the handbike, but that it out-performs lever-propelled and push-rim wheelchairs, suggesting that this novel design is more suited to indoor use and may therefore be an attractive alternative to push-rims for activities of daily living.

Key words: wheelchair; upper extremity; metabolism; rehabilitation; activities of daily living.

 

 

Manual wheelchair propulsion is the most favoured mode of propulsion adopted by a large percentage of wheelchair users: More than 90% of all self-propelled wheelchairs are ambulated via push-rims by using the arms to apply force (1). Push-rim propulsion (PRP) is energetically inefficient, highly strenuous for the musculoskeletal system, and associated with high cardiopulmonary effort (1–6). Furthermore, PRP often leads to severe upper limb injuries, especially at the shoulder and wrist joints (7–9).

Gross mechanical efficiency (GME), defined as the percentage ratio between external power output (PO) and energy expenditure (En) is often used to benchmark mechanical efficiency of manual wheelchair propulsion. Oxygen uptake per unit time (VO₂), heart rate (HR) and propulsion frequency are the parameters typically used to assess the metabolic cost and efficiency of wheelchair propulsion (10, 11). Due to high physical strain during PRP most of the expended energy dissipates, for example in heat loss, the rest contributes to propulsion. Thus, GME is typically measured to be in the range 2–10% and rarely exceeds 10% (1, 12). Despite similar power output conditions, reported values for GME vary widely, which may be explained by individual differences in the physical ability of wheelchair users and by the influence of propulsion speed and surface properties (1, 12).

Alternative modes of manual wheelchair propulsion have been tested, the most common alternatives being lever-propulsion and arm-crank propulsion. Compared with PRP, both of these methods increase the joint range of motion in the upper limb, particularly at the wrist and shoulder joints (1, 5, 11). Lever-propelled devices were mainly designed to reduce repetitive strain injuries (1, 6, 11, 13). In general, GME in lever-propelled devices is reported to be higher compared with PRP, and wheelchair users report greater overall satisfaction with lever-propelled wheelchairs, but previous designs often do not consider user anthropometrics (1, 4, 13). Handbikes are the most popular arm-crank-propelled alternative to PRP for manual wheelchair propulsion, with values for GME reported to range from 8% to 15% (5, 14–17). Due to a higher energetic efficiency and lower strain on the cardiorespiratory system, several investigators have recommended the handbike as an alternative to push-rim wheelchair propulsion for outdoor use (5, 14–16, 18–20). Although the efficiency of alternative devices for wheelchair propulsion is often higher, most of these are limited to outdoor use because they are bulky, heavier, less convenient for transferring, and less manoeuvrable (1, 5, 15).

Our group has developed a novel handle-based propulsion (HBP) mechanism for conventional wheelchairs as a compact indoor alternative to PRP (21). With an ergonomically optimized propulsion movement and the ability to continuously apply propulsive force, HBP offers a continuous cyclic motion at ergonomic joint ranges of motion and has been shown to decrease joint excursions and maximum joint torques during propulsion (21–23). The aim of the current study was to investigate differences in metabolic cost and mechanical efficiency for this novel HBP device compared with the standard PRP. Data were collected from paraplegic subjects, who are long-term wheelchair users, and non-disabled individuals, to further investigate effects on propulsion mode and efficiency and to determine how the results differ for long-term wheelchair users who have trained muscle coordination patterns for PRP and may also have changed relative muscle strengths. Each group used an instrumented wheelchair-based test rig operating at constant speed and different resistance levels. It was hypothesized that, under similar conditions, HPB would be more energetically efficient and less strenuous than conventional PRP, due to continuous force application.

METHODS

Subjects

Eight right-handed paraplegic (PP) subjects were recruited from the spinal cord injury (SCI) rehabilitation centre “Weisser Hof”, in Klosterneuburg, Austria. PP subjects were eligible for the study if they had an SCI level between L1 and T12, no permanent medication and no history of upper-limb injury.

The 10 non-disabled (ND) students (controls), who participated in this study (Table I), were eligible if they were right-handed, had no history of upper limb injury, and no permanent medication. The significantly younger ND group had no experience with PRP, whereas all PP subjects were long-term wheelchair users with a minimum experience of 3 years. The sample size was defined by the maximum number of PP subjects available, rather than based on statistical considerations. All subjects provided informed consent and approval for the study was obtained from the responsible federal state ethics committee Ethikkommission für das Land Niederösterreich (NÖ Ethikkommission), Landhausplatz 1, Haus 15B, 3109 St. Pölten, Austria (GS1-EK-3/149-2018).

Table I. Characteristics of paraplegic (PP) and non-disabled (ND) subjects

Subject (sex)	Age, years	Weight, kg	Height, cm	BMI, kg/m²	SCI level
PP
1 (f)	54	62	179	19	T11–12
2 (m)	27	80	188	23	T6–7
3 (f)	43	75	160	29	T10–L1
4 (f)	56	63	158	25	T12–L1
5 (m)	52	65	175	21	T12–L2
6 (m)	45	85	192	23	T11–L1
7 (m)	21	62	185	18	T8–9
8 (m)	51	93	173	31	T12–L2
Mean (SD)	44 (12)	73 (11)	176 (12)	24 (5)
ND
1 (f)	23	70	165	26
2 (f)	21	63	170	22
3 (m)	21	92	185	27
4 (f)	21	54	162	21
5 (m)	23	58	181	18
6 (m)	24	70	169	25
7 (f)	26	58	175	19
8 (m)	36	99	181	30
9 (f)	35	64	176	21
10 (m)	19	61	187	17
Mean (SD)	25 (5)	69 (14)	175 (8)	22 (4)
SD: standard deviation; SCI: spinal cord injury; BMI: body mass index; m: male; f: female.

Experimental set-up

All subjects were tested using a previously developed test rig (24) (Fig. 1) consisting of a lightweight manual wheelchair (Eurochair Vario,XXL, Meyra Orthopedics, Germany, Kalletal) with 0.42 m seat depth, 0.50 m seat width, and an adjusted seat height of 0.51m. The wheelchair was mounted on a square tube frame to avoid direct contact between the wheels and the floor and to facilitate mounting of different hand-driven propulsion devices. A controlled brushless motor combined with gearbox and flywheel mounted under the seat of the wheelchair provided torques that simulated resistances during wheelchair propulsion. Timing belts were used to promote a slip-free power transmission to the wheels and front connection points. For this study, the test rig operated in maximal power mode, where the resistance progressively increased according to a predefined resistance increment and time interval. In all trials, visual feedback allowed the participant to maintain the target speed during propulsion.

Fig. 1. Wheelchair-based test rig with mounted handle-based propulsion (HBP) devices: 1: wheelchair; 2: HBP devices; 3: mounting frame; a: front attachment pulley; b: back wheel hub; c: rear attachment pulley; d: resistance power unit.

Propulsion devices

For PRP, conventional 24-inch (609,6 mm) diameter push-rim wheels were mounted on the test rig whereas for crank propulsion 2 HBP devices (25) were utilized instead of the armrests (Figs 1 and 2). Each HBP device consisted of a rotating crank on which a handle was mounted. During propulsion, a sliding guide changed the length of the crank during each rotation, allowing the handle to follow the optimized propulsion path (22). The gear ratio of the HBP device was fixed at 1.2.

Fig. 2. Handle-based propulsion (HBP) device and its components. The horizontal and vertical positions of the crank centre can be adjusted according to the user’s body anthropometry.

Test protocol

All participants performed the same exercise with both PRP and HBP consecutively on the same day, with a least 10 min rest between exercises. The subjects were instructed to refrain from smoking for at least 2 h before testing, to not consume any caffeinated or alcoholic beverages, and to void their bladder shortly before the measurements. To control the influence of fatigue on the effect of mode, subjects were assigned to alternately start with PRP or HBP, i.e. even-numbered subjects started the exercises with PRP, and odd-numbered subjects with HBP. To ensure familiarization with all equipment, subjects participated in a short preliminary session in which both HBP and PRP were used with low resistance (5 W). The exercise test consisted of 2 parts: a 2-min submaximal exercise test performed at 15 W constant resistance power, followed by a maximal exercise test in which the resistance power was increased by 5 W every minute. Both parts were performed consecutively with no break in between. The test was terminated when either 55 W of resistance power was reached or when the subject reached physical exhaustion and could not continue. To simulate a common propulsion speed, subjects were asked to maintain a propulsion speed in the range 1.20–1.65 m/s for both HBP and PRP, similar to the range of speeds used in previous studies (14, 26, 18). Sixteen of the 18 subjects completed all exercises. One female participant was excluded due to problems with balance during testing, while 1 male ND subject was excluded because the required speed range was not achieved during the experiments.

Data collection

Actual resistance power (PO, W) and linear velocity (v, m/s) were measured concurrently during each test. In addition, oxygen uptake (VO₂, mL/min), carbon dioxide output (VCO₂, mL/min) and ventilation (Ve, mL/min) were measured continuously using a wearable metabolic measurement system (Cosmed K5, Cosmed GmbH, Fridolfing, Germany), while heart rate (HR, beats/min) was measured using a mobile chest heart rate monitor (Polar H10 ANT+, Polar Electro Inc., Kempele, Finland). The spirometry system was matched to the subjects with respect to ethnicity (Caucasian) and calibrated with a reference gas after each subject. Linear velocity (v, m/s) and cadence (RPM, 1/min) were obtained from the test rig control and were also measured using cadence sensors (B00JLMS848 ANT+/B00JLMRXCQ ANT+, Garmin Ltd., Schaffhausen, Switzerland) mounted on both HPB cranks and the back wheels.

Data analysis

Respiratory exchange ratio (RER, –) was calculated as the ratio between carbon dioxide output (VCO₂, mL/min) and oxygen uptake (VO₂, mL/min). Metabolic expenditure (En, W) was obtained from VO₂ and RER using the equation reported by Garby et al. (27):

Gross mechanical efficiency (GME, %) was defined as the ratio between the actual resistance power (PO, W) provided by the test rig and energy expenditure (En) of the subject:

Weight-specific oxygen uptake (VO₂/kg, mL/min/kg) was obtained from oxygen uptake (VO₂) and the weight of the subject. Oxygen uptake per unit distance travelled (VO₂ efficiency, mL/kg/m) was found from the measured VO₂/kg and the corresponding mean linear velocity (v).

Mean submaximal values of VO₂, VO₂/kg, VO₂ efficiency, VCO₂, Ve, RER, HR and GME for both HBP and PRP were measured during the last minute of the 2-min (15 W) submaximal exercise. During the maximal power exercise, the subject’s mean values of VO₂, VO₂/kg, VO₂ efficiency, VCO₂, Ve, RER, HR and GME were calculated for both propulsion modes at each achieved resistance level to provide a comparison between the 2 propulsion modes. Peak values of VO₂, VO₂/kg, VO₂ efficiency, VCO₂, Ve, RER were found by calculating the highest mean value of each variable measured over a time-interval of 30 s, whereas the peak value of HR was defined as the highest mean value of HR measured over an interval of 10 s. Peak power output was defined as the highest resistance level achieved during the maximal exercise, which was maintained for at least 30 s.

Statistical analysis

All analyses were performed using SPSS (IBM SPSS Statistics 26, SPSS, Inc., Chicago, IL, USA). Mean and peak values were calculated using descriptive statistics. An analysis of variance (ANOVA) with a 2 × 2 design (mode: HBP, PRP; group: ND, PP) was used to determine the effect of propulsion mode and subject group on submaximal exercise responses. An ANOVA with a 2 × 4 × 2 design (mode: HBP, PRP; power output: 15, 20, 25 and 35 W; group: ND, PP) was also used to evaluate the interaction between propulsion mode, power output, and subject group for the resistance levels achieved by all subjects during the maximal exercise tests. The effect of propulsion mode and subject group on peak performance was found using an ANOVA with a 2 × 2 design (mode: HBP, PRP; group: ND, PP). Statistical significance for all tests was set at p < 0.05 with no adjustment for multiple comparisons.

RESULTS

Submaximal exercise

There was a significant main effect of propulsion mode on all parameters except HR, indicating lower VO₂, VO₂/kg, Ve, RER and higher GME and VO₂ efficiency during HBP (Table II). Mean values of VO₂ efficiency during HBP were 0.07 mL/kg/m lower for ND and 0.03 mL/kg/m lower for PP. Similarly, mean GME was 1.03% higher for ND and 2.75% higher for PP during HBP compared with PRP. There was a significant main effect of group on all parameters except Ve and HR, indicating that the mean values were significantly different between ND and PP. HR in both subject groups did not show any significant effects. A significant interaction effect was found for mode and group on RER, indicating that the changes in RER during the HBP and PRP tests were different between PP and ND.

Table II. Mean (standard deviation) submaximal values measured for handle-based propulsion (HBP) and push-rim propulsion (PRP) at 15W constant resistance for the paraplegic (PP) and non-disabled (ND) groups

	GME, %	VO2 efficiency, mL/kg/m	VO2/kg, mL/min/kg	VO2, mL/min	Ve, L/min	RER –	HR, bpm
ND, n	9	9	9	9	9	9	9
HBP	7.59 (1.37)	0.12 (0.02)	9.94 (2.28)	620.04 (106.17)	20.42 (4.24)	0.83 (0.06)	110.00 (17.84)
PRP	6.56 (1.73)	0.19 (0.06)	10.94 (2.82)	738.93 (198.00)	24.74 (7.23)	0.87 (0.06)	108.80 (21.26)
PP, n	7	7	7	7	7	7	7
HBP	9.9 1(1.41)	0.08 (0.02)	6.72 (1.71)	486.72 (68.19)	17.35 (2.41)	0.68 (0.05)	96.10 (11.14)
PRP	7.16 (1.22)	0.11 (0.02)	8.89 (1.97)	649.70 (104.18)	23.76 (4.87)	0.84 (0.05)	111.33 (11.59)
Mode
F-value	13.23	11.38	3.80	8.82	8.60	26.37	1.41
p-value	0.001	0.002	0.049	0.006	0.007	< 0.001	0.245
Group
F-value	7.92	17.68	10.48	5.50	1.22	21.07	0.92
p-value	0.009	< 0.001	0.003	0.026	0.279	< 0.001	0.345
Mode × Group
F-value	2.70	1.52	0.52	0.22	0.33	7.78	1.93
P-value	0.111	0.229	0.478	0.646	0.573	0.009	0.176
GME: gross mechanical efficiency; VO2 efficiency: oxygen uptake per unit distance travelled; VO2/kg: weight-related oxygen uptake; VO2: oxygen uptake; Ve: ventilation, RER: respiratory exchange ratio; HR: heart rate; bpm: beats per min. Bold values indicate significance (p < 0.05).

Maximal exercise

Maximum power resistance levels were different between PP and ND. Both groups achieved higher power levels for HBP than PRP. Values for GME were higher in HBP than PRP for both the PP and ND groups (Fig. 3). Furthermore, HR increased as resistance level increased. A significant main effect of propulsion mode for all parameters in HBP was evident, indicating higher GME and VO₂ efficiency and lower VO₂/kg, VO₂, Ve, HR and En (Table III). HR was significantly higher for the PP group during PRP, whereas no significant differences were found with respect to propulsion mode for the ND group. Except for HR, there was also a significant main effect of group for all outcome variables, indicating higher GME and VO₂ efficiency and lower physiological responses for the PP group compared with the ND group. A significant interaction effect between propulsion mode and group was found for VO₂ efficiency and HR, suggesting that the differences between HBP and PRP were larger for VO₂ efficiency and smaller for HR in the ND group compared with the PP group. No interaction effects were found between propulsion mode, group, and power output, indicating that the combined effect of power output and propulsion mode was similar for both groups. In both groups, mean RER values were lower for HBP than PRP (Fig. 4). As power resistance increased, RER values were above 1.0 for PRP but remained below 1.0 for HBP, indicating higher physical exhaustion during PRP.

Table III. Analysis of variance (ANOVA) results obtained for the maximal exercise tests (15–35 W) for the paraplegic (PP) and non-disabled (ND) subject groups

	Mode		Group		Mode group		Mode PO		Mode group PO
	F-value	p-value	F-value	p-value	F-value	p-value	F-value	p-value	F-value	p-value
GME	68.410	< 0.001	31.782	< 0.001	3.127	0.079	0.049	0.995	0.386	0.818
VO2 efficiency	37.194	< 0.001	26.753	< 0.001	9.375	0.003	2.584	0.040	0.676	0.610
VO2/kg	24.859	< 0.001	39.676	< 0.001	0.098	0.755	0.668	0.615	0.400	0.808
VO2	56.350	< 0.001	24.057	< 0.001	0.023	0.879	0.641	0.634	0.083	0.988
Ve	53.835	< 0.001	10.167	0.002	0.049	0.825	1.968	0.103	0.042	0.997
HR	18.579	< 0.001	1.955	0.164	12.999	< 0.001	0.626	0.645	0.048	0.996
En	62.511	< 0.001	30.945	< 0.001	0.000	0.995	0.839	0.503	0.098	0.983
GME: gross mechanical efficiency; VO2 efficiency: oxygen uptake per distance travelled; VO2 /kg: weight related oxygen uptake; VO2: oxygen uptake; Ve: ventilation; HR: heart rate; EN: energy expenditure; PO: power output (resistance level). Bold values indicate significance (p < 0.05).

Fig. 3. Dependence of gross mechanical efficiency (GME), oxygen uptake per unit distance travelled (VO₂ efficiency), weight-related oxygen uptake (VO₂/kg), heart rate (HR), ventilation (Ve) and oxygen uptake (VO₂) on resistance level measured for paraplegic (PP) subjects (orange lines, triangle markers) and non-disabled (ND) subjects (blue lines, circle markers) during handle-based propulsion (HBP) (solid line) and push-rim propulsion (PRP) (dashed lines).

Fig. 4. Mean respiratory exchange ratio (RER) values and standard deviations measured for handle-based propulsion (HBP) (solid shading) and push-rim propulsion (PRP) (diagonal stripes shading) in paraplegic (PP, orange) and non-disabled (ND, blue) subjects.

Peak responses

A significant main effect was observed for propulsion mode on PO_(peak), Ve_(peak) and HR_(peak); specifically, power output was higher and Ve and HR were lower for HBP than PRP (Table IV). No main effect was found for propulsion mode on peak oxygen uptake (VO₂/kg_(peak), VO_2(peak)) and RER_(peak). There was a significant main effect on group for all parameters except HR_(peak). However, no interaction effect was found for propulsion mode and group on any of the outcome variables, as the ND and PP groups showed similar trends in peak values during HBP and PRP.

Table IV. Mean values and standard deviations for peak power output (PO_(peak)), peak weight-related oxygen uptake (VO₂/kg_(peak)), peak oxygen uptake (VO_2(peak)), peak ventilation (Ve_(peak)), peak respiratory exchange ratio (RER_(peak)) and peak heart rate (HR_(peak))

	PO(peak), W	VO2/kg(peak), mL/min/kg	VO2(peak), mL/min	Ve(peak), L/min	RER(peak) –	HR(peak, bpm
ND, n	9	9	9	9	9	9
HBP	55.56 (7.68)	37.86 (30.70)	1,615.42 (236.50)	59.69 (8.59)	1.09 (0.07)	157.67 (23.01)
PRP	46.67 (6.61)	34.41 (8.85)	1,833.95 (541.13)	77.58 (26.71)	1.15 (0.09)	161.89 (21.12)
PP, n	7	7	7	7	7	7
HBP	46.43 (4.76)	19.42 (2.01)	1,351.39 (178.50)	43.17 (7.05)	0.99 (0.06)	134.86 (7.13)
	42.86 (4.88)	25.19 (8.13)	1,541.17 (182.56)	57.79 (9.20)	1.04 (0.07)	158.86 (12.31)
Mode
F-value	7.78	0.03	2.89	8.20	3.87	4.87
p-value	0.009	0.856	0.100	0.008	0.059	0.036
Group
F-value	8.38	4.90	5.37	10.23	13.81	4.08
p-value	0.007	0.035	0.028	0.003	0.001	0.053
Mode × Group
F-value	1.42	0.54	0.01	0.08	0.02	2.39
p-value	0.244	0.468	0.906	0.776	0.888	0.133
Results of the analysis of variance (ANOVA) for propulsion mode (handle-based propulsion (HBP) vs push-rim propulsion (PRP)), group (ND vs PP) and interaction effects are also shown. PO: power output; bpm: beats per minute. Bold values indicate significance (p < 0.05).

DISCUSSION

The aim of this study was to measure metabolic cost and mechanical efficiency during wheelchair propulsion using a novel HBP device and to compare these performance indicators against those measured for conventional PRP. The study investigated physiological parameters for both propulsion modes in a combined submaximal and maximal exercise test using both PP and non-disabled (ND) subjects. The results of the submaximal test showed that higher GME and VO₂ efficiency were attained with lower physiological responses during HBP compared with PRP. This effect was also observed during the maximal exercise tests with continuously increasing resistance (i.e. increasing power output).

Furthermore, HBP showed higher peak power output and lower peak heart rate compared with PRP. In all tests, subjects showed significantly higher efficiency and lower physiological responses during HBP compared with PRP. Thus, our hypothesis that HPB is more efficient and less strenuous than conventional PRP was supported.

The results for the submaximal and maximal tests showed higher GME and lower physiological responses with the HBP device, which is consistent with findings from previous studies that have focused on comparing submaximal arm-crank exercise with conventional PRP under similar conditions (2, 14, 28, 29). Our measured values of GME and VO₂ efficiency for PRP are also consistent with data reported previously by others (10, 30). The higher GME and VO₂ efficiency achieved by the PP group compared with the ND control group may be due to their familiarity with PRP as well as better upper-limb muscle conditioning resulting from everyday use of the wheelchair. However, the results of the maximal exercise indicate that the effects of propulsion mode and power output apply to both groups. Previous studies on arm-crank devices report higher peak power output compared with conventional PRP (14, 29). Similar findings were observed during the propulsion mode with HBP, where mean peak power output values were 8.3% higher for the PP group and 19.0% higher for the ND group compared with PRP. Regarding peak oxygen uptake, we found significant differences between the subject groups but no significance was found between HBP and PRP, which is consistent with the results of other studies comparing arm-crank devices with PRP (14, 31–34).

Particularly for subjects with paraplegia, mean heart rates at power levels between 15W and 35W as well as HR_(peak) were significantly reduced using the HBP device, which is also supported by lower RER values. This suggests a higher endurance capacity of the paraplegic subjects, which is also reflected in higher mean peak power output values achieved in HBP compared with PRP.

However, comparing the differences in HR values between both subject groups, no statistical significance was found: an observation that has also been reported previously and has been attributed to the normally higher HR in paraplegic persons (14).

Thus, independent of subject group, HBP was shown to achieve higher mechanical efficiencies and lower oxygen uptake at both submaximal and maximal workloads compared with PRP.

Although studies reported in the literature on alternative propulsion devices are not easily compared, the data on efficiency presented here, particularly our measurements of GME and VO₂ efficiency, place the HBP below the performance of the handcycle but above lever-propelled and PRP wheelchairs. Moreover, the current findings indicate that propelling a wheelchair with this novel HBP device is more efficient and less strenuous than propelling a wheelchair with a push-rim. In addition, our previous study (23) showed that the HBP device may lower the probability of upper limb injuries by reducing joint loads and ergonomic joint ranges. To adapt to an individual user’s wheelchair, the horizontal and vertical adjustment mechanisms shown in Fig. 2 can be eliminated, leading to an even more compact and lightweight design of the crank mechanism that can also be swivelled for transfers. Therefore, we consider the novel wheelchair drive to be more suitable for indoor use than, for example, handcycles, making it an attractive alternative to push-rims for activities of daily living.

This study has some limitations. The sample size (PP n = 8; ND n = 10) was relatively small. Because force application during wheelchair propulsion is highly individualized, especially for PRP, a larger number of test subjects may influence the findings reported here. In addition, the age difference between the 2 groups may have affected the group comparison, as, in general, younger subjects are more likely to have higher cardiorespiratory fitness. However, it is generally difficult to establish groups of able-bodied and paraplegic subjects that are comparable in terms of cardiorespiratory fitness, as it has been well investigated (35) that paraplegic subjects have lower cardiorespiratory fitness in comparison with able-bodied subjects. In any case, only group comparisons may have been influenced by the age difference, the effects of propulsion mode are not affected.

Another improvement that is recommended for further studies is to make the test rig adjustable in the medial-lateral direction to also account for individual differences in body anthropometry in this dimension.

In conclusion, we found that both subject groups demonstrated significantly higher peak power output, higher mechanical efficiency, and lower physiological responses during HBP compared with PRP. The results indicate that propelling the wheelchair with this novel HBP device is less strenuous and more efficient than conventional push-rim propulsion. Overall, the performance of the HBP was below that of the handbike, but exceeded the performances of the lever-propelled and push‑rim wheelchairs.

ACKNOWLEDGEMENTS

This work was supported by the Austrian Science Fund (FWF) under Grant P 25507-B24.

The authors acknowledge TU Wien Bibliothek for financial support through its Open Access Funding Programme.

REFERENCES

1. van der Woude LH, Dallmeijer AJ, Janssen TW, Veeger D. Alternative modes of manual wheelchair ambulation: an overview. Am J Phys Med Rehabil 2001; 80: 765–777. DOI: 10.1097/00002060-200110000-00012
2. van der Woude LH, GROOT G de, Hollander AP, van Ingen Schenau GJ, ROZENDAL RH. Wheelchair ergonomics and physiological testing of prototypes. Ergonomics 1986; 29: 1561–1573. DOI: 10.1080/00140138608967269
3. van der Woude LH, Veeger HE, ROZENDAL RH, Sargeant AJ. Optimum cycle frequencies in hand-rim wheelchair propulsion. Wheelchair propulsion technique. Eur J Appl Physiol Occup Physiol 1989; 58: 625–632. DOI: 10.1007/BF00418509
4. Lui J, MacGillivray MK, Sheel AW, Jeyasurya J, Sadeghi M, Sawatzky BJ. Mechanical efficiency of two commercial lever-propulsion mechanisms for manual wheelchair locomotion. J Rehabil Res Dev 2013; 50: 1363–1372. DOI: 10.1682/JRRD.2013.02.0034
5. Flemmer CL, Flemmer RC. A review of manual wheelchairs. Disabil Rehabil Assist Technol 2016; 11: 177–187. DOI: 10.3109/17483107.2015.1099747
6. Requejo PS, Lee SE, Mulroy SJ, Haubert LL, Bontrager EL, Gronley JK et al. Shoulder muscular demand during lever-activated vs pushrim wheelchair propulsion in persons with spinal cord injury. J Spinal Cord Med 2008; 31: 568–577. DOI: 10.1080/10790268.2008.11754604
7. Samuelsson KAM, Tropp H, Gerdle B. Shoulder pain and its consequences in paraplegic spinal cord-injured, wheelchair users. Spinal Cord 2004; 42: 41–46. DOI: 10.1038/sj.sc.3101490
8. Wei S, Huang S-L, Jiang C-J, Chiu J-C. Wrist kinematic characterization of wheelchair propulsion in various seating positions: implication to wrist pain. Clin Biomech (Bristol, Avon) 2003; 18: 46-52. DOI: 10.1016/s0268-0033(03)00084-6
9. Rao SS, Bontrager EL, Gronley JK, Newsam CJ, Perry J. Three-dimensional kinematics of wheelchair propulsion. IEEE Trans Rehabil Eng 1996; 4: 152–160. DOI: 10.1109/86.536770
10. Beekman CE, Miller-Porter L, Schoneberger M. Energy cost of propulsion in standard and ultralight wheelchairs in people with spinal cord injuries. Phys Ther 1999; 79: 146–158. DOI: 10.1093/ptj/79.2.146
11. Zukowski LA, Roper JA, Shechtman O, Otzel DM, Bouwkamp J, Tillman MD. Comparison of metabolic cost, performance, and efficiency of propulsion using an ergonomic hand drive mechanism and a conventional manual wheelchair. Arch Phys Med Rehabil 2014; 95: 546–551. DOI: 10.1016/j.apmr.2013.08.238
12. Hopman MT, van Teeffelen WM, Brouwer J, Houtman S, Binkhorst RA. Physiological responses to asynchronous and synchronous arm-cranking exercise. Eur J Appl Physiol Occup Physiol 1995; 72: 111–114. DOI: 10.1007/BF00964124
13. Rifai Sarraj A, Massarelli R, Rigal F, Moussa E, Jacob C, Fazah A et al. Evaluation of a wheelchair prototype with non-conventional, manual propulsion. Ann Phys Rehabil Med 2010; 53: 105–117. DOI: 10.1016/j.rehab.2009.12.001
14. Dallmeijer AJ, Zentgraaff IDB, Zijp NI, van der WOUDE LHV. Submaximal physical strain and peak performance in handcycling versus handrim wheelchair propulsion. Spinal Cord 2004; 42: 91–98. DOI: 10.1038/sj.sc.3101566
15. van der Woude LHV, Groot S de, JANSSEN TWJ. Manual wheelchairs: research and innovation in rehabilitation, sports, daily life and health. Med Eng Phys 2006; 28: 905–915. DOI: 10.1016/j.medengphy.2005.12.001
16. Verellen J, Theisen D, Vanlandewijck Y. Influence of crank rate in hand cycling. Med Sci Sports Exerc 2004; 36: 1826–1831. DOI: 10.1249/01.mss.0000142367.04918.5a
17. Kraaijenbrink C, Vegter RJK, Hensen AHR, Wagner H, van der Woude LHV. Different cadences and resistances in sub-maximal synchronous handcycling in able-bodied men: effects on efficiency and force application. PLoS One 2017; 12: e0183502. DOI: 10.1371/journal.pone.0183502
18. Arnet U, van Drongelen S, Veeger DH, van der Woude L HV. Force application during handcycling and handrim wheelchair propulsion: an initial comparison. J Appl Biomech 2013; 29: 687–695. DOI: 10.1123/jab.29.6.687
19. van der Woude LH, Bosmans I, Bervoets B, Veeger HE. Handcycling: different modes and gearratios. J Med Eng Technol 2000; 24: 242–249. DOI: 10.1080/030919000300037168
20. Dallmeijer AJ, Ottjes L, Waardt E de, van der WOUDE LHV. A physiological comparison of synchronous and asynchronous hand cycling. Int J Sports Med 2004; 25: 622–626. DOI: 10.1055/s-2004-817879
21. Babu Rajendra Kurup N, Puchinger M, Gföhler M. Forward dynamic optimization of handle path and muscle activity for handle based isokinetic wheelchair propulsion: a simulation study. Comput Methods Biomech Biomed Engin 2019; 22: 55–63. DOI: 10.1080/10255842.2018.1527321
22. Kurup NBR, Puchinger M, Keck T, Gfoehler M. Wrist kinematics and kinetics during wheelchair propulsion with a novel handle-based propulsion mechanism. Annu Int Conf IEEE Eng Med Biol Soc 2018; 2018: 2146–2149. DOI: 10.1109/EMBC.2018.8512658
23. Puchinger M, Stefanek P, Gstaltner K, Pandy MG, Gfohler M. In vivo biomechanical assessment of a novel handle-based wheelchair drive. IEEE Trans Neural Syst Rehabil Eng 2021; 29: 1669–1678. DOI: 10.1109/TNSRE.2021.3105388
24. Puchinger M, Kurup NBR, Gföhler M. A Test rig for investigating manual wheelchair propulsion devices: TAR 2017 - Technically Assisted Rehabilitation - March 09-10, 2017. Current Directions in Biomedical Engineering 2017; 3: 12. DOI: 10.1515/cdbme-2017-1001
25. Babu Rajendra Kurup N, Puchinger M, Gfoehler M. A preliminary muscle activity analysis: handle based and push-rim wheelchair propulsion. J Biomech 2019; 89: 119–122. DOI: 10.1016/j.jbiomech.2019.04.011
26. Valent LJM, Dallmeijer AJ, Houdijk H, Slootman HJ, Janssen TW, Post MWM, et al. Effects of hand cycle training on physical capacity in individuals with tetraplegia: a clinical trial. Phys Ther 2009; 89: 1051–1060. DOI: 10.2522/ptj.20080340
27. Garby L, Astrup A. The relationship between the respiratory quotient and the energy equivalent of oxygen during simultaneous glucose and lipid oxidation and lipogenesis. Acta Physiol Scand 1987; 129: 443–444. DOI: 10.1111/j.1365-201X.1987.tb10613.x
28. Sawka MN, Glaser RM, Wilde SW, Luhrte TC von. Metabolic and circulatory responses to wheelchair and arm crank exercise. J Appl Physiol Respir Environ Exerc Physiol 1980; 49: 784–788. DOI: 10.1152/jappl.1980.49.5.784
29. Tropp H, Samuelsson K, Jorfeldt L. Power output for wheelchair driving on a treadmill compared with arm crank ergometry. Br J Sports Med 1997; 31: 41–44. DOI: 10.1136/bjsm.31.1.41
30. Groot S de, Dallmeijer AJ, Kilkens OJ, van Asbeck FW, Nene AV, Angenot EL, et al. Course of gross mechanical efficiency in handrim wheelchair propulsion during rehabilitation of people with spinal cord injury: a prospective cohort study. Arch Phys Med Rehabil 2005; 86: 1452–1460. DOI: 10.1016/j.apmr.2004.11.025
31. Glaser RM, Sawka MN, Brune MF, Wilde SW. Physiological responses to maximal effort wheelchair and arm crank ergometry. J Appl Physiol Respir Environ Exerc Physiol 1980; 48: 1060–1064. DOI: 10.1152/jappl.1980.48.6.1060
32. McConnell TJ, Horvat MA, Beutel-Horvat TA, Golding LA. Arm crank versus wheelchair treadmill ergometry to evaluate the performance of paraplegics. Paraplegia 1989; 27: 307–313. DOI: 10.1038/sc.1989.46
33. Martel G, Noreau L, Jobin J. Physiological responses to maximal exercise on arm cranking and wheelchair ergometer with paraplegics. Paraplegia 1991; 29: 447–456. DOI: 10.1038/sc.1991.61
34. Gass EM, Harvey LA, Gass GC. Maximal physiological responses during arm cranking and treadmill wheelchair propulsion in T4-T6 paraplegic men. Paraplegia 1995; 33: 267–270. DOI: 10.1038/sc.1995.60
35. Farkas GJ, Gordon PS, Swartz AM, Berg AS, Gater DR. Influence of mid and low paraplegia on cardiorespiratory fitness and energy expenditure. Spinal Cord Ser Cases 2020; 6: 1–10. DOI: 10.1038/s41394-020-00363-5