Introduction
The sport of people with disabilities has gained increased attention, recognition and acceptance in media and society. The process of professionalisation can be seen in the development of attendance, performance, and scientific exami-nations. Because of the highly individual properties, a huge potential for the paralympic sport is lying in the field of sports biomechanics [44]. As a part of paracycling, handcycling became a very common sport for people with spinal cord injuries or amputations on recreational und international level. An optimi-sation in handcycling propulsion kinematics and kinetics could improve the performance of interational athletes.
Purpose
The aim of the current study was to examine the kinematics and kinetics of handcycling propulsion during an incremental test. In addition, parameters that are crucial for the sportspecific performance were identified.
Methods
Twelve non-disabled male triathletes (26.0±4.4 yrs., 1,83±0.06 m, 74.3±3.6 kg) without handcycling experience perfromed an initial familarization followed by a 15-s All-Out test. The tests were performed in a racing handcycle (Shark S, Sopur, Sunrisemedical, Malsch, Germany) that was attached to a ergometer (Cyclus 2, 8 Hz, RBM elektronik-automation GmbH, Leipzig, Germany). Out of this sprint test, the peak power output (POmax,AO15) and the glycolytic rate (V̇Lamax) were determined. The incremental test started with an initial load of 20 W and increased every 5 min by 20 W until subjective exhaustion. In the end of every stage, the rate of perceived exertion (RPE) on a global (cardio-pulmo-nary) and local (upper extremity) level, arterialized capillary blood lactate concentration of the ear lobe (La [mmol l-1]) and heart rate (HR [min-1]) were recorded. Performance criteria were the maximal power output during the incremental test (POmax,ST) and the calculated lactate threshold based on the fixed 4 mmol l-1 method (PO4mmol) [32; 4]. The kinematic and kinetic measure-ments occured at the end of the first and beginning of the last minute of each stage for 20 seconds. Seven high speed infrared cameras (100 Hz, MX-F40 and MX-3+, Vicon Nexus 2.3, Vicon Motion Systems Ltd., Oxford, UK) were placed aroud the handcycle. 44 spherical retro-reflective Markers were placed on the crank, the ergometer and anatomical landmarks according to the UpperLimb-Model of Vicon Nexus. The joint kinematics considered the angles and angular velocities of shoulder-flexion (SF), shourler-abduction (SA), shoulder-rotation (SR), elbow-flexion (EF), palmar-flexion (PF), radial-duction (RD) und trunk-flexion (RF). The tangential crank kinetics based on a power-meter (1000 Hz, Schobener Bike Management System, SRM, Jülich, Ger-many) installed in the crank. Out of the crank angular velocity and torque sig-nals, the acute power output (PO) was calculated. The data were averaged, filtered (4th-order low-pass Butterworth filter, cut-off frequency 10 Hz) and resampled to a length of 360 frames per cycle using MATLAB (R2016a, MathWorks®, Natick, Massachusetts, USA). Parameters of kinematics and kinetics were maximum and minimum value (MinV and MinI), range (MaxV-MinV), the crank angle of the maximum and minimum (MaxI and MinI) and the values at maximal and minimal acute PO (@MaxPO and @MinPO). Additionally, the accomplished work within one cycle during six sectors (Press-down, 330 to 30°; Pull-down, 30 to 90°; Pull-up, 90 to 150°; Lift-up, 150 to 210°; Push-up, 210 to 270°; Push-down, 270 to 330°) was calculated. Changes within the incremental test were analysed usind a one-way ANOVA with repeated measures [28]. In case of missing assumption of normal distribution (Kolmogorov-Smirnov test with Lilliefors-correction), the non-parametric Fried-man test was applied. Post-Hoc comparisons based on Bonferroni. The calculation of partial eta-squared (ηp2) was added as effect-size. To identify determinants of performance, bivariate Pearson’s correlation coefficient was calculated. For parameters with significant difference to normal distribution, the non-parametric correlation coefficient of Spearman was applied. The level of significance was set to α = 0.05 [28]. For specific comparisons of means, the effect size Cohen’s d was calculated [18].
Results
The mean POmax,AO15, V̇Lamax, POmax,ST and PO4mmol were 545±70 W, 0.44±0.11 mmol l-1 s-1, 131±15 W, 87±12 W. RPElocal at exhaustion was significantly higher than RPEglobal (p = 0.003, d = 2.03). The maximal torque was found within the Pull-down or Push-up sector, whereas the lowest torque and ca-dence occured during the Lift-up. During the incremantal test, a significant decrease in retroversion (p < 0.0005, ηp2 = 0.346), adduction (p < 0.0005, ηp2 = 0.527), elbow-flexion (p < 0.0005, ηp2 = 0.572) and elbow-extension (p < 0.0005, ηp2 = 0.658) was observed. Maximal abduction (p < 0.0005, ηp2 = 0.723) and internal rotation (p = 0.031, ηp2 = 0.046) showed an increase. The MaxV-MinV of the SF (p = 0.069, ηp2 = 0.252), EF (p < 0.0005, ηp2 = 0.775) and RD (p < 0.0005, ηp2 = 0.438) rather decreased, whereas a rather increase in MaxV-MinV for SA (p = 0.003, ηp2 = 0.410), SR (p = 0.069, ηp2 = 0.035) and RF (p = 0.009, ηp2 = 0.385) was found. The maximal elbow-flexion (p = 0.002, ηp2 = 0.414), elbow-extension (p = 0.006, ηp2 = 0.310) and dorsal-flexion (p < 0.0005, ηp2 = 0.356) occured significantly later in crank cycle. At the 180° position, a signifiantly higher abduction (p < 0.0005, ηp2 = 0.405) and lower elbow-flexion (p = 0.001, ηp2 = 0.202) was measured.
The angular velocity of all degrees of freedom (df) increased during the incremental test. The maximal angular velocity of anteversion (p = 0.004, ηp2 = 0.052), abduction (p = 0.009, ηp2 = 0.410), dorsal-flexion (p = 0.053, ηp2 = 0.211), ulnar-duction (p = 0.037, ηp2 = 0.169) and trunk-flexion (p = 0.006, ηp2 = 0.092) occured later in crank cycle. The Lift-up was performed with a significantly higher anteversion velocity (p < 0.0005, ηp2 = 0.348), internal-rotation velocity (p = 0.006, ηp2 = 0.092) and dorsal-flexion velocity (p = 0.024, ηp2 = 0.0284). The proportion of work showed a decrease of the Press-down from 20 W to 80 (p = 0.021, d = -1.22) and 100 W (p = 0.041, d = -1.11) and an increase of the Pull-down from 20 to 120 W (p = 0.011, d = 1.33).
POmax,ST significantly correlated with Lamax,ST (p = 0.015, r = -0.680), V̇Lamax (p = 0.022, r = -0.649), and PO4mmol (p = 0.050, r = 0.577). The variability of cadence was negatively correlated with POmax,ST (p = 0.009, r = -0.711). Near the 180° position, a lower internal-rotation (p = 0.019, r = -0.662), higher elbow flexion (p = 0.005, r = 0.750), and lower dorsal-flexion (p = 0.003, r = 0.775) was beneficial for POmax,ST. An early maximal external-rotation velocity (p = 0.080, r = -0.511) and abduction velocity (p = 0.079, r = 0.525) and a late dorsal-flexion velocity (p = 0.213, p = -388) tended to result in higher POmax,ST.
For PO4mmol, a significant correlation was found to the training load of the participants [h w-1] (p = 0.029, r = 0.628), and the maximal lactate concen-tration during the incremental test (p = 0.037, r = -0.605). Participants with a high MaxV-MinV in SF (p = 0.016, r = 0.676) and RD (p = 0.023, r = -0.492) and a low MaxV-MinV in EF (p = 0.087, r = -0.514) and PF (p = 0.461, r = -0.231) achieved higher PO4mmol. At 180°, a lower internal-rotation (p = 0.012, r = -0.698) and dorsal-flexion (p = 0.226, r = 0.349) and a higher abduction (p = 0.174, r = 0.420) and elbow-flexion (p = 0.214, 0.387) was beneficial for PO4mmol. The maximal anteversion velocity positively correlated with PO4mmol (p = 0.027, r = 0.349). A late occurrence of the maximal anteversion velocity (p = 0.023, r = 0.646), adduction velocity (p = 0.039, r = 0.600), dorsal-flexion velocity (p = 0.031, r = 0.623), and trunk-flexion velocity (p = 0.048, r = 0.579) and early occurrence of the maximal elbow-extension velocity (p = 0.018, r = -0.665) resulted in higher PO4mmol. At 180°, a high maximal abduction velocity (p = 0.026, r = 0.636), dorsal-flexion velocity (p = 0.003, r = -0.774), and radial-duction velocity (p = 0.025, r = 0.640) resulted in higher PO4mmol.
Discussion
The increase in cadence was higher than the increase in torque, which is consistent with literature, assuming that high muscle forces and concomitant limitations in local blood flow are responsible for exhaustion in arm cranking exercises [59; 66]. Another incidence for an especially local based fatigue, are the significantly higher RPE values on local level. The kinematic and kinetic results indicate that the maintenance of high PO is primaliry limited by the clean like motion near the 180° crank angle, which defines the transition between pull and push phase. For maximal PO, a reinforced pull phase could narrow the loss in crank angular velocity during the clean and thus delay fatigue. Participants who perform the clean with a higher retroversion and lower abduction and internal-rotation are advantaged. For submaximal PO, the change in direction of the force vector (especially during the clean) should be as tangential as possible to avoid unnecessary work. An active wrist motion in RD could improve the propulsion economy.
Transfered to practice, a shorter crank arm and slim backrest seem to be suitible alterations of the handcycle setting. A sport specific strength training of the upper extermity is important to improve perfromance. Especially the wrist, cheast and shoulder muscles should be trained in strength and endu-rance.
Limitations of the study mosty refer to the unexperienced and non-disabled participants. Therefore, future studies should replicate the current study with elite handcyclists and expand the test spectrum by sprint and continous loads considering the examination of muscle activation patterns (MAP).