We present a protocol to test the aerobic and anaerobic power of the upper body muscles over a duration of 3 min in able-bodied as well as in paraplegic and tetraplegic individuals. The protocol presents specific modifications in its application for upper-body exercise in individuals with or without disability.
Reliable exercise protocols are required to test changes in exercise performance in elite athletes. Performance improvements in these athletes may be small; therefore, sensitive tools are fundamental to exercise physiology. There are currently many exercise tests that allow for the examination of exercise capacity in able-bodied athletes, with protocols mainly for lower-body or whole-body exercise. There is a trend to test athletes in a sport-specific setting that closely resembles the actions that the participants are used to performing. Only a few protocols test short-term, high-intensity exercise capacity in participants with an impairment of the lower body. Most of these protocols are very sport-specific and are not applicable to a wide range of athletes. One well-known test protocol is the 30 s Wingate test, which is well-established in cycling and in arm crank exercise testing. This test analyzes high-intensity exercise performance over a 30 s time duration. In order to monitor exercise performance over a longer duration, a different method was modified for application to the upper body. The 3 min, all-out arm crank ergometer test allows athletes to be tested in a manner specific to 1,500 m wheelchair racing (in terms of exercise duration), as well as to upper body exercises such as rowing or hand-cycling. In order to increase the reliability with identical test conditions, it is crucial to precisely replicate settings such as the resistance (i.e., torque factor) and the position of the participants (i.e., the height of the crank, the distance between the crank and the participant, and the fixation of the participant). Another important issue concerns the beginning of the exercise test. Fixed revolutions per minute are required to standardize the test conditions for the start of the exercise test. This exercise protocol shows the importance of accurate operations to reproduce identical test conditions and settings.
There are several exercise tests that accurately determine the increase in exercise performance in elite athletes over the course of a training period1,2,3,4,5. One of these tests is the reliable 3-min all-out exercise test on a braked cycling ergometer3,4,5,6. This test was used to determine critical power, but it was also applied to exercise testing with athletes, as well as to research7,8,9. As this test was mainly used for lower-extremity performance, such as in rowing7 and cycling3,5, a similar testing protocol for upper-body exercise was needed. Sport disciplines that mainly use the upper body might be possible beneficiaries for such a new test protocol, in addition to athletes or individuals with an impairment of lower body muscles (e.g., an amputation or an impairment of limbs due to a spinal cord injury). Hence, a test protocol on the arm crank ergometer is a good tool to easily test upper-body exercise performance in a variety of athletes from different sport disciplines.
The existence of a very similar 30 s Wingate arm crank ergometer test10,11 helped with the development of a protocol for a 3 min, all-out arm crank ergometer test. Its duration is very similar to that of a 1,500 m wheelchair race. Therefore, this new test protocol of the 3 min, all-out arm crank ergometer test was tested for its test-retest reliability12. Overall, the reliability of this test protocol was excellent, so it could be a future testing tool in the field of upper-body exercise testing. Nevertheless, the use of this exercise test requires attention, especially when testing individuals with a spinal cord injury. Therefore, the aim of this experimental article is to demonstrate a detailed protocol that describes not only the test settings and analysis of the test results, but that also indicates the differences between testing able-bodied individuals and athletes with a spinal cord injury.
The study was approved by the local ethical committee (Ethikkommission Nordwest- und Zentralschweiz, Basel, Switzerland), and written informed consent was obtained from the participants before starting the study.
1. Test Preparation and Participant Instruction
2. Execution of the Exercise Protocol
3. Data Analysis and Interpretation of the Results
The test-retest reliability was checked in 21 recreationally trained (but not specifically upper-body trained), non-smoking individuals (9 males, 12 females; age: 34 ± 11 years; body mass: 69.6 ± 11.1 kg; and height: 175.5 ± 6.9 cm). Table 1 shows the results for the relative and absolute test-retest reliabilities12. The peak power compared between the test and retest is presented in Figure 112. A Bland Altman plot for this test-retest is presented in Figure 212. Afterwards, this 3-min, all-out arm crank ergometer test was used in 17 able-bodied (age: 38 ± 7 years, height: 183 ± 13 cm, and body mass: 79 ± 6 kg), 10 paraplegic, and 7 tetraplegic participants (Table 2). Individual data representing an able-bodied as well as a tetraplegic participant are presented in Figure 3. Able-bodied participants showed a peak power of 483 ± 94 W, whereas paraplegic and tetraplegic participants were found to have a peak power of 375 ± 101 W and 98 ± 49 W, respectively. The mean power was found to be 172 ± 20 W, 157 ± 28 W, and 40 ± 14 W for able-bodied, paraplegic, and tetraplegic participants, respectively. Significant differences in mean and peak power were found between able-bodied and tetraplegic participants (p < 0.001), as well as between paraplegic and tetraplegic participants (p < 0.001). The end lactate concentration was 8.9 ± 2.4 mM in able-bodied participants, 10.6 ± 2.9 mM in paraplegic participants, and 4.0 ± 0.8 mM in tetraplegic participants. The average heartrate during the 3-min all-out test was 155 ± 9.2 bpm in able-bodied, 163 ± 6.2 bpm in paraplegic, and 113 ± 15.9 bpm in tetraplegic participants. Again, the heartrate of tetraplegic participants was significantly lower compared to paraplegic (p <0.001) as well as to able-bodied participants (p <0.001). The oxygen consumption measured during a 3-min all-out test is presented in Figure 4.
Figure 1: Mean power compared between the two 3-min all-out exercise tests on an arm crank ergometer12. The solid line represents the best fit and the dashed line the line of identity. Please click here to view a larger version of this figure.
Figure 2: Bland-Altman plot for peak power12. SD = standard deviation. Please click here to view a larger version of this figure.
Figure 3: Individual data of the 3-min all-out arm crank ergometer test for an able-bodied as well as a tetraplegic participant. Left = able-bodied participant; right = tetraplegic participant; blue line = power output; green line = cadence. Please click here to view a larger version of this figure.
Figure 4: Oxygen consumption during a 3 min all-out arm crank ergometer test in an able-bodied participant. Time point zero represents the start of the 3 min test. The data are presented as the raw data measured breath-by-breath.
ICC | 95% CI | SEM % | SRD % | CV | |
Peak power [W] | 0.961 | [0.907; 0.984] | 2 | 5.6 | 6.66 |
Mean power [W] | 0.984 | [0.960; 0.993] | 0.6 | 1.6 | 3.13 |
Minimal power [W] | 0.964 | [0.914; 0.985] | 1.4 | 4 | 6.05 |
Time to peak [s] | 0.379 | [-0.052; 0.691] | 22.5 | 62.4 | 11.37 |
Fatigue index | 0.940 | [0.858; 0.975] | 3.6 | 9.9 | 9.43 |
Rel. peak power [W/kg] | 0.922 | [0.818; 0.968] | 2.8 | 7.8 | 6.45 |
Rel. mean power [W/kg] | 0.950 | [0.882; 0.979] | 1.1 | 3.2 | 3.46 |
Total work [J] | 0.984 | [0.960; 0.993] | 0.6 | 1.6 | 3.13 |
Table 1: Test-retest reliability for all parameters12. ICC = intra-class correlation coefficient; CI = confident interval; SEM = standard error of the measurement; SRD = smallest real difference; CV = coefficient of variation; rel. = relative.
Paraplegic participant | Lesion level | AIS | Age (y) | Body mass (kg) | Height (cm) | ||
P01 | Th12 | A | 47 | 80 | 184 | ||
P02 | Th10 | A | 43 | 73 | 183 | ||
P03 | Th11 | A | 55 | 72 | 174 | ||
P04 | L1 | A | 26 | 64 | 150 | ||
P05 | Th12 | A | 22 | 63 | 185 | ||
P06 | L1 | A | 32 | 76 | 175 | ||
P07 | Th11 | A | 59 | 80 | 178 | ||
P08 | L1 | A | 35 | 63 | 165 | ||
P09 | L4 | A | 44 | 78 | 176 | ||
P10 | L1 | A | 48 | 80 | 185 | ||
Mean | 41 | 73 | 176 | ||||
SD | 12.1 | 6.8 | 10.4 | ||||
Tetraplegic participant | Lesion level | AIS | Age (y) | Body mass (kg) | Height (cm) | ||
T01 | C5 | A | 24 | 85 | 188 | ||
T02 | C7 | A | 31 | 60 | 180 | ||
T03 | C7 | A | 40 | 60 | 168 | ||
T04 | C7 | A | 31 | 80 | 190 | ||
T05 | C5 | A | 43 | 80 | 176 | ||
T06 | C6 | A | 56 | 74 | 170 | ||
T07 | C5 | A | 65 | 75 | 190 | ||
Mean | 41 | 73 | 180 | ||||
SD | 13.6 | 9.9 | 9.3 |
Table 2: Anthropometric data of paraplegic and tetraplegic participants. AIS = American Spinal Injury Association Impairment Scale, Th = thoracic, L = lumbar, C = cervical, SD = standard deviation.
Exercise testing in spinal-cord injured athletes is crucial to tracking exercise performance over several months or years of training. Only a few exercise tests exist to check short-term, high-intensity exercise performance on the arm crank ergometer. This method describes in detail how an exercise test that was already examined for its reliability in cycling5 and rowing7 might be applied to the arm crank ergometer. To collect reliable and meaningful results, two factors are very important: first, the preparation of the participant for this exercise test and second, the standardization of the test. Thus, the participant is instructed to enter the lab in a rested state, meaning no intense training during the two days prior to the test. Food intake (e.g., carbohydrate-rich nutrition 24 h prior testing) and sleep (e.g., at least 7 h of sleep during the two nights prior to testing) also need to be taken into account. Additionally, before performing the first "real" exercise test, a familiarization trial should be performed to ensure that the participant understands the test protocol. The second condition concerns the standardization of the test protocol, which includes the warmup, the start of the test, and the fixation strategies for the wheelchair and hands (step 1.2). These settings need to be kept identical for each test with the same person. Also, the height of the crank may influence the power output and oxygen consumption20. Additionally, the handgrip position might be clinically relevant in terms of shoulder pain following the exercise21. Furthermore, abdominal binding might influence respiratory function and oxygen transport22, but it seems to augment trunk stability and exercise performance23. Thus, it is best to record adjustments and fixations in detail to reproduce equivalent conditions in the following test.
The results in Figures 1 and 2 indicated that this 3 min, all-out arm crank test is reliable in able-bodied participants and can be used for research or exercise testing12. When comparing able-bodied, paraplegic, and tetraplegic participants with each other, differences between these three groups were found. Able-bodied and paraplegic participants showed very similar results for maximal and mean power, whereas tetraplegic participants performed with a significantly lower power output. Similar findings were shown when comparing the 30 s Wingate test for paraplegic participants11 with the one for tetraplegic participants10. Due to the different lesion levels of these spinal-cord injured participants, various muscles are affected by the impairment. Thus, with a higher lesion level (e.g., in tetraplegic participants), a less active muscle mass results in a lower power output (Table 2). To diminish the variability of the lesion, only individuals with a lesion level between cervical 5 (C5) and 7 (C7) and with a motor and sensory complete spinal cord injury were included to represent tetraplegic participants. Nevertheless, even between such participants, a high inter-individual variability can occur. Individuals with an injury below C5 only have Musculus (M.) biceps activated in the arms, whereas individuals with an injury below C7 show activity in M. biceps brachii, M. extensor radialis, and M. triceps arm muscles24. Therefore, even with very narrow inclusion criteria like the ones mentioned here, the group of participants was very inhomogeneous in terms of muscle function. Concerning paraplegic and able-bodied participants, one would expect a higher power output in able-bodied participants due to full trunk stability and unaffected respiratory muscle function. Our results did not show any significant differences in power output between these two groups, although able-bodied participants performed slightly better. This may have been because our able-bodied study participants were less trained compared to the paraplegic participants. It is possible that, even with lower trunk stability and lower respiratory function, the paraplegic participants were more adapted to arm cranking, which might have compensated for their disadvantages.
Even though this test protocol seems to be reliable in able-bodied individuals, test-retest reliability might be limited in tetraplegic individuals. The power output in tetraplegic participants showed a very high inter-individual variability, which might be explained by differences in muscle function and power (e.g., active muscle dependent on lesion level). Nevertheless, Jacobs, Johnson, Somarriba, and Carter showed a very high reliability in the shorter version of this test (30 s Wingate test) in tetraplegic athletes10. We assumed that the reliability might be good enough over a longer duration (i.e., 3 min in this protocol) but did not test it. The lack of a test-retest reliability investigation in individuals with a tetraplegia is a limitation. Thus, before using this test protocol in a group of individuals with tetraplegia, we recommend checking the reliability of the protocol first. Another limitation of the study involves the standardization of the fixation strategies in participants with a very low core stability (e.g., individuals with a high lesion level, such as in tetraplegia). To fix them to the chair, a strap must be bound around the chair and the participant. Thus, it can be be difficult to record the height of the strap and how tightly it was pulled. In a future study, it would be beneficial to test whether the tightness of such a strap influences the power output in the test protocol.
Nevertheless, this test is a good tool to test athletes from different sport disciplines involving upper-body exercise. Additionally, as the pacing strategy of going all-out right at the beginning of the test is predefined, no individual pacing strategy is present.
To conclude, this test seems to be a reliable tool to test exercise performance over a duration of 3 min and is similar to short-term, high-intensity exercise performance. A familiarization trial is needed before its application in athletes or in research investigations in order to minimize learning effects from test 1 to test 2. In addition, further research is needed to check test-retest reliability in participants or athletes with a tetraplegia.
The authors have nothing to disclose.
We are grateful for the assistance from Martina Lienert and Fabienne Schaufelberger during exercise testing, as well as from P. D. Claudio Perret, PhD for his scientific advice.
Angio V2 arm crank ergometer | Lode BV, Groningen, NL | N/A | arm crank ergometer |
Lode Ergometry Manager Software | Lode BV, Groningen, NL | N/A | Software |
10ul end-to-end capillary | EKF-diagnostics GmbH, Barleben, Germany | 0209-0100-005 | Capillaries |
haemolysis cup | EKF-diagnostics GmbH, Barleben, Germany | 0209-0100-006 | hemolysis cup |
lactate analyzer | Biosen C line, EKF-diagnostics GmbH | 5213-0051-6200 | lactate analyzer |
Heart rate monitor, Polar 610i | Polar, Kempele, Finland | P610i | heart rate monitor |
metabolic cart, Oxygen Pro | Jaeger GmbH | N/A | metabolic cart |
oxygen mask, Hans Rudolph | Hans Rudolph Inc. , USA | 113814 | oxygen mask |
statistical software, PSAW Software | SPSS Inc., Chicago USA | N/A | statistical software |
desinfectant, Soft-Zellin | Hartmann GmbH, Austria | 999979 | desinfectant |
Quality control cup, EasyCon Norm | EKF-diagnostics GmbH, Barleben, Germany | 0201-005.012P6 | quality control |
Quality control cup 3mmol/L | EKF-diagnostics GmbH, Barleben, Germany | 5130-6152 | control cup |
Chip sensor lactate analyzer | EKF-diagnostics GmbH, Barleben, Germany | 5206-3029 | chip sensor |
Lactate system solution | EKF-diagnostics GmbH, Barleben, Germany | 0201-0002-025 | lactate system solution |
lancet, Mediware Blutlanzetten | medilab | 54041 | lancet |
Calibration gas, | Jaeger GmbH | 36-MC G020 | calibration gas |
chair provided by distributor (ergoselect) | ergoline GmbH, Germany | N/A | chair provided by distributor |