Analysis of spatial and temporal step parameters during crutch-assisted gait as a dual-task: a pilot study

Journal of Biomechanical Engineering 1 Corresponding author: Carmen Ridao-Fernández Analysis of spatial and temporal step parameters 1 during crutch-assisted gait as a dual-task: a pilot study 2 3 Carmen Ridao-Fernández MCs (Ridao-Fernández C)1 4 Department of Physiotherapy, Research Group “Area of Physiotherapy CTS-305”, 5 University of Seville, Seville, Spain. 6 Email address: [email protected] 7 8 Joaquín Ojeda PhD (Ojeda J) 9 Department of Mechanical Engineering and Manufacture, Research Group "Mechanical 10 Engineering”, University of Seville, Seville, Spain. 11 Email address: [email protected] 12 13 Gema Chamorro-Moriana PhD (Chamorro-Moriana G) 14 Department of Physiotherapy, Research Group “Area of Physiotherapy CTS-305”, 15 University of Seville, Seville, Spain. 16 Email address: [email protected] 17 18 19 Journal of Biomechanical Engineering 2 Corresponding author: Carmen Ridao-Fernández 1Corresponding author information: 20 Postal address: C/ Avicena s/n. C.P. 41009. Sevilla. España. 21 Telephone (business): 0034695417193 22 Journal of Biomechanical Engineering 3 Corresponding author: Carmen Ridao-Fernández Analysis of spatial and temporal step parameters 23 during crutch-assisted gait as a dual-task: a pilot study 24 25 Abstract 26 27 The main objective was to analyze the changes in the spatial and temporal step 28 parameters during a dual-task: walking with a forearm crutch to partially unload the 29 body weight of the subject. The secondary objective was to determine the influence of 30 the use of the crutch with the dominant or non-dominant hand in the essential gait 31 parameters. Seven healthy subjects performed gait without crutches (GWC) and 32 unilateral assisted gait (UAG) with the crutch carried out by dominant hand (DC) and 33 non-dominant hand (NDC). Gait was recorded using a Vicon System; the GCH System 34 2.0 and the GCH Control Software 1.0 controlled the loads. The variables were: step 35 length, step period, velocity, step width and step angle. The Wilcoxon signed-rank test 36 compared GWC and UAG while also analyzing the parameters measured for both legs 37 with DC and NDC in general and in each subject. Wilcoxon test only found significant 38 differences in 1 of the 15 general comparisons between both legs. In the analysis by 39 subject, step length, step period and velocity showed significant differences between 40 GWC and UAG. These parameters obtained less differences in DC. The effect of a 41 forearm crutch on UAG caused a reduction in step length and velocity, and an increase 42 in step period. However, it did not entail changes in step angle and step width. UAG 43 Journal of Biomechanical Engineering 4 Corresponding author: Carmen Ridao-Fernández was more effective when the dominant hand carried the crutch. The unloading of 10% 44 body weight produced an assisted gait which closely matched GWC. 45 Journal of Biomechanical Engineering 5 Corresponding author: Carmen Ridao-Fernández Introduction 46 47 Gait is one of the main functions of the human being [1]. That is why many 48 researchers try to analyze it from many perspectives. Firstly, biomechanics analyzes gait 49 parameters in detail. Secondly, innovative technology develops more effective and 50 efficient objective evaluation methods to improve the accuracy of quantitative data 51 [2,3]. Thirdly, the clinical area requires appropriate action protocols or clinical 52 procedures to re-educate gait when there is an alteration caused by a pathology or 53 another triggering factor such as the help systems employed [3]. This article aims to 54 bring together the biomechanical, technological and clinical vision. 55 56 This study focuses on assisted gait with Canadian or forearm crutches. They are 57 frequently employed in the recovery of musculoskeletal injuries of the lower limb 58 which require partial weight unloading [4]. This gait modality involves a dual-task for 59 the subject since the crutch is a distracting effect against the automatic movement of 60 normal gait [5,6]. The dual task paradigm establishes that the introduction of a second 61 task during a cognitive or motor performance leads to a possible competition between 62 the attentional resources available [7,8]. The tasks usually cited in literature are 63 cognitive, which have been extensively analyzed. On the other hand, motor dual-tasks 64 have been recently studied by authors such as Santos [9], Brustio [8], who analyzed 65 walking while performing a second manual task such as buttoning or unbuttoning a 66 shirt. Assisted gait had previously been considered a dual-task [6], although it has not 67 been studied until now. The second task increases activity in every case at the prefrontal 68 Journal of Biomechanical Engineering 6 Corresponding author: Carmen Ridao-Fernández cortex, and performance in either task decreases [10]. The attentional demands of a task 69 and the interference effects of current task could be influenced by several factors, such 70 as the nature of the tasks involved [10], the level of skill and the dominance in 71 unilateral tasks [11]. In the latter, the dominant hand has a lower coordinating 72 requirement than non-dominant hand [11]. Thus, it is easier for the subject to perform 73 assisted gait with a single crutch when this assistive device is carried out by the 74 dominant hand rather than non-dominant. Besides, the level of coordination of the 75 subject is essential in a dual-task. Consequently, this study must analyze healthy 76 subjects so as to avoid the interference of the pathology in the task [12]. 77 78 The step is the reference unit in this cyclic activity. Moreover, gait can be 79 asymmetrical, even in people who do not suffer pathologies [13]. Therefore, this article 80 studies the parameters related to step and its symmetry objectively and in detail: length, 81 time, width and angle along with gait velocity. These parameters are essential, so they 82 are part of many functional gait assessment scales that are relevant in the clinical 83 setting, such as Gait and Balance Scale [2], Tinetti Mobility Test [14] and Chamorro 84 Assisted Gait Scale [15]. This kinematic analysis of gait will allow patients to reduce 85 energy expenditure during its execution [16] and consequently to optimize human gait. 86 Besides, it will contribute to increasing the scientific evidence of clinical actions [17]. 87 In this way, new treatment guidelines can be established to re-educate gait, initially 88 assisted by crutches and progressing to free gait [4]. 89 90 The main objective of this study is to analyze the changes in the spatial and 91 temporal step parameters of gait during a dual-task: walking with a forearm crutch to 92 Journal of Biomechanical Engineering 7 Corresponding author: Carmen Ridao-Fernández partially unload the body weight of the healthy subject. The secondary objective of this 93 article is to determine the influence of the use of the crutch with the dominant or non-94 dominant hand in the essential parameters of gait. Regarding the study objectives, the 95 hypotheses of the authors were two. First, the incorporation of a forearm crutch on 96 assisted gait leads to a change in step parameters (especially in step length, step period 97 and gait velocity). Secondly, assisted gait is more correct, that is, more closely 98 matching gait without crutches when the crutch is carried by dominant hand than non-99 dominant hand. 100 Journal of Biomechanical Engineering 8 Corresponding author: Carmen Ridao-Fernández Methods 101 102 Participants 103 104 The sample was selected in a non-probabilistic and convenience sampling. It 105 consisted of 7 healthy Caucasian subjects from south-west Europe (3 men and 4 106 women), with an intermediate socio-cultural and socio-economic status. All subjects 107 were right-handed. The mean age was 30.3 years (SD=10.39), with a minimum of 21 108 and a maximum of 53. 109 110 The inclusion criteria were: aged between 18 and 60 years old; absence of any 111 musculoskeletal or neurological injury, previous experience with crutches (to have used 112 crutches due to a musculoskeletal injury) and normal gait (asymptomatic and without 113 any observable alteration in the gait pattern with free cadence on walking) [18]. As 114 additional inclusion criterium, the subjects were asked to pass a simple static 115 equilibrium test which consisted of maintaining monopodal balance for 30 seconds on 116 each foot without suffering any great bodily movements (overacted movement, mainly 117 arms, time and a motor skill to rebalance) [19]. 118 119 The exclusion criterion was: evident general coordination and physical ability 120 disorder that could alter the normal or assisted gait, such as a history of vestibular or 121 Journal of Biomechanical Engineering 9 Corresponding author: Carmen Ridao-Fernández neurological disorders, proprioceptive musculoskeletal alterations or cerebellar tumors, 122 among others. 123 124 All participants signed a consent form in accordance with the Helsinki 125 Declaration. This study was approved by the Ethics Committee of the Hospital 126 University Virgen Macarena, Seville, Spain. 127 128 Data collection 129 130 In this study, the Vicon® System (Oxford, UK) of three-dimensional motion 131 analysis (3D) [20], instrumented crutches denominated GCH 2.0 Load Measurement 132 System (University of Seville, Seville, Spain) for assisted gait [21], and the GCH 133 Control Software 1.0 program (University of Seville, Seville, Spain) [22] were used. 134 135 Data collection was carried out under laboratory conditions (with the same 136 artificial light and temperature) on an 8.5-meter-long walkway. The subjects walked 137 with a 2-point assisted gait, with simultaneous heel and crutch support. The selected 138 height of the crutch corresponded to an elbow flexion of 20 ° to 30 ° [23]. The required 139 velocity was at free cadence. The participants completed the walk on 10 occasions for 140 each gait modality. First, they performed gait without crutches (GWC) and then 141 unilateral assisted gait (UAG): with the crutch carried in the dominant hand, hereafter 142 Journal of Biomechanical Engineering 16 Corresponding author: Carmen Ridao-Fernández Table 3 Step length analysis 254 255 Table 4 Step period analysis 256 257 Table 5 Velocity analysis 258 259 Table 6 Step angle analysis 260 261 Table 7 Step width analysis 262 263 Step length maintained steady values in the descriptive analysis (means 264 oscillated between 0.53m and 0.74m) and obtained 20/28 significant confidence 265 intervals at a 99% confidence level (Table 3). In addition, 15 of these 20 significant 266 values of p showed a high effect size (>0.8). Regarding step period, means had a broad 267 range (52s to 86s). Besides the fact that 27/28 results had significant results at a 99% 268 confidence level, 26 of them obtained a high effect size (Table 4). Gait velocity, whose 269 means oscillated between 0.63 and 1.27m/s, took bigger values for GWC than UAG in 6 270 subjects. This parameter showed 12/14 significant results at a 99% confidence level, all 271 of which with a high effect size (Table 5). Step length had less significant differences in 272 DC than NDC (Table 3), which also came up in step period and velocity. Step angle and 273 step width, nevertheless, obtained only 14/28 and 3/14 significant results respectively 274 (Tables 6 and 7). 275 Journal of Biomechanical Engineering 17 Corresponding author: Carmen Ridao-Fernández 276 Gait symmetry 277 278 The comparison of the variables measured for right and left legs was performed 279 using Wilcoxon signed-rank test (p=0.05). Right and left step parameters were 280 compared both in GWC and UAG. Significant differences were found in only one of the 281 15 comparisons: GWC step period (Left and Right). 282 283 Age, height and weight 284 285 The Pearson correlation coefficients, calculated in relation to age, height and 286 weight of subjects, did not find significant results. 287 Journal of Biomechanical Engineering 18 Corresponding author: Carmen Ridao-Fernández Discussion 288 289 This study analyzed the effect that a forearm crutch has on gait. The load exerted 290 was minimal and comfortable for the subject, approximately 10% of body weight. This 291 amount of load was required so as not to distort the correct gesture of gait, that is, the 292 biomechanically correct gait pattern developed during GWC. Other authors have stated 293 that alterations on gait appear when the unloading of body weight on a single crutch is 294 more than 25% [27]. Despite this, the authors of this research defend that, even with 295 percentages below 25% of unloading, there is an alteration in the body position and gait 296 parameters that impedes its functional development. Therefore, a lower load was 297 required to guarantee a correct gesture of gait. 298 299 Several points were analyzed in this work. On one hand, the main spatial and 300 temporal parameters in relation to the step were studied by including the distracting 301 effect of the crutch. On the other hand, the study considered whether the crutch was 302 carried in the dominant or non-dominant hand. Thus, the influence of coordination on 303 this dual-task was studied without interference from excessive lower limb unloading for 304 a single crutch; and neither from pathology, as it was a sample from healthy subjects. 305 306 Crutch-assisted gait as a dual-task 307 Assisted gait was analyzed from the dual-task paradigm [7]. The use of forearm 308 crutches implies a distracting effect for gait. Therefore, it is a more complex task, which 309 requires the attention of the subject [6,28]. Thus, the level of coordination of the latter 310 Journal of Biomechanical Engineering 19 Corresponding author: Carmen Ridao-Fernández acquires special relevance. Other authors, such as Kelly [29], Woollacott [30] and 311 Simoni [10], have studied the influence of velocity, step period, and step length on gait 312 when another simultaneous task was performed. Despite the widely analyzed second 313 tasks during gait being cognitive (as saying the days of the week backwards [29] or 314 pronouncing alternate letters of the alphabet [10]), different motor dual-tasks have been 315 investigated recently. That is, walking while performing a second manual-task as 316 screwing and unscrewing a bottle cap in front of the body [8]. Gait while carrying a 317 forearm crutch has not been analyzed from this paradigm until now. However, assisted 318 gait has previously been considered a dual-task [5,6]. Moreover, the influence of the 319 forearm crutch has not been analyzed when carried in the dominant or non-dominant 320 hand within this kind of study. 321 322 Step length 323 Regarding the results, step length did not show significant changes in the overall 324 analysis (Table 3 2). However, the study of this variable allowed us to observe 325 modifications when considering the gait progress of each subject. This main parameter 326 of the step remains unchanged in individuals considered as coordinated. This was the 327 case of subject 6, where there was no difference of means between GWC and UAG. The 328 level of coordination decreased the cognitive demands of gait and, therefore, the subject 329 could satisfy both tasks simultaneously [6]. However, this parameter showed variations 330 resulting from the need to adapt to the use of the crutch when coordination was 331 insufficient to perform the dual-task effectively. There was a reduction in step length as 332 is reported in the literature [31]. Even so, this parameter was the best adapted parameter 333 to the demands of gait, as it showed less significant differences between GWC and 334 Journal of Biomechanical Engineering 20 Corresponding author: Carmen Ridao-Fernández UAG than step period or velocity. These results were supported by the overall analysis 335 (Table 2). Furthermore, a slight difference was observed between comparisons of DC 336 and NDC in Table 3. Less significant differences were found when the subjects carried 337 the crutch in their dominant hand. The authors justify this fact as all the subjects were 338 right-handed [11]. In addition, this also occurred in the step period and velocity. 339 340 Step period and gait velocity 341 The change in step length was accompanied by a reduction in velocity and an 342 increase in time as a consequence of using the crutch. These outcomes coincided with 343 those by Santos et al. [9], in their study about motor dual-task. They concluded that 344 walking while performing a second manual task produced a deficit in the second task. 345 Regarding the study results, more significant differences were obtained in the step 346 period than step length, which were also reinforced by high effect sizes. The subjects 347 showed less ability in maintaining this parameter despite adding the crutch to gait and 348 changing it by hand. For both steps and both crutches, step period was modified by 349 increasing the load exerted, controlled by the GCH 2.0 Load Measurement System. This 350 evidence prompted us to analyze the effect of the loads prospectively in detail, as is 351 shown at the end of this section. 352 353 Step angle and step width 354 The step angle did not follow a characteristic pattern in relation to the studied 355 variables. According to the different results shown in Table 6, the step angle was not 356 related to the changes that occurred in gait when performing the dual-task of UAG. The 357 Journal of Biomechanical Engineering 21 Corresponding author: Carmen Ridao-Fernández step width was not modified either by including the crutch effect as indicated in Table 8 358 7. An increase in the step width could be expected to increase the base of support as a 359 result of the coordinating needs of the dual-task [33]. However, the authors thought that 360 it would be in cases of greater instability in which an increase would be observed in the 361 base of support, as frequently happens in neurological patients [34]. 362 363 Gait symmetry 364 The analysis of the parameters of the right step regarding the left when carrying 365 a forearm crutch allowed the study of asymmetries. As was shown in the results section, 366 14 of the 15 comparisons made between the right and left leg showed no significant 367 differences. In these cases, both steps were similar. The addition of a simultaneous task 368 to the gait did not alter the gait of the subjects in terms of step asymmetry. This was due 369 to the level of coordination and learning that the subjects had [6]. In addition, they were 370 all healthy, and the crutch load was minimal. These are ideal conditions for a functional, 371 optimum and symmetrical gait. Despite these optimal study conditions, a clinical 372 tendency to asymmetry was observed. Subjects tended to increase the time and reduce 373 the length and velocity of the contralateral step to the crutch when it was carried in the 374 non-dominant hand [31]. Therefore, dominance is a factor to be borne in mind in gait 375 re-education in the clinical setting [35]. The physiotherapist should pay more attention 376 to the learning of UAG when the non-dominant hand carries the crutch. It should be 377 done especially in patients with a reduced level of coordination, either due to their own 378 characteristics or as a result of the proprioceptive affection derived from the pathology 379 [36]. The dual-task learning process will allow them to master the task, even to the point 380 of automation. That is, to do it without paying attention. The improvement of motor 381 Journal of Biomechanical Engineering 22 Corresponding author: Carmen Ridao-Fernández dual-task performance caused by training was recently demonstrated by Brustio et al. 382 [8]. Besides, a good learning will prevent an incorrect gesture or limp (due to 383 asymmetry in the step length, arrhythmic steps, etc.), remain in time even when the 384 patient does not need to use the crutch. In this way, correct re-education of assisted gait 385 will be achieved, also optimizing treatment time as well as avoiding relapses of the 386 injury [15]. 387 388 389 390 391 Study limitations 392 393 Even though it is an innovative and complex study due to using different 394 technologies, the sample size was limited. The authors propose to study the assisted gait 395 prospectively as a dual-task on a higher number of healthy subjects to confirm the 396 results obtained in this pilot study. Although the 65% power obtained was interpreted as 397 a good result, with a 95% confidence level and a sample composed of 7 subjects, XX 398 subjects will be necessary to reach 95% power in future work. 399 400 Journal of Biomechanical Engineering 23 Corresponding author: Carmen Ridao-Fernández In addition, a study on the behavior of gait parameters is being developed in 401 which different percentages of body weight are unloaded. That is why authors observed 402 an increase in assisted gait failure on increasing the load. The results of both studies will 403 be complemented in the analysis of assisted gait with a forearm crutch. 404 405 Finally, it is necessary to study the trunk parameters (deviation of the center of 406 gravity, pelvic dissociation, etc.) and upper limb in the UAG, in order to know the 407 asymmetries that occur in said body areas [37]. 408 Journal of Biomechanical Engineering 24 Corresponding author: Carmen Ridao-Fernández Conclusions 409 410 This study showed that the effect of the forearm crutch on UAG as a dual-task 411 results in a reduction in step length, an increase in step period and a reduction in 412 velocity. The angle and width of the step were not modified by the influence of the 413 crutch in this dual-task. These results confirmed the hypotheses established. In addition, 414 this article demonstrated that the dominance of the subject influences the UAG. As was 415 expected by the authors, it is more effective when the crutch is carried in the dominant 416 hand. The UAG was symmetrical, because the subjects were coordinated and performed 417 a minimum weight loss in the absence of a pathology. The unloading of 10% body 418 weight of the subject allowed a functional gait from the biomechanical point of view, 419 which closely matched gait without crutches. Despite the symmetry, a tendency to 420 asymmetry between both steps was observed when the non-dominant hand carried the 421 crutch, which was significant from the clinical point of view. Dominance will therefore 422 be a factor to be kept in mind by the physiotherapist. They should pay more attention to 423 gait re-education when the non-dominant hand carries the crutch, and especially when 424 the subject coordination level is reduced. All this will allow to make the gesture 425 automatic properly and, thus, the time of treatment will be optimized, so avoiding 426 relapses in the injury. 427 Journal of Biomechanical Engineering 25 Corresponding author: Carmen Ridao-Fernández Conflict of interest statement 428 429 None of the authors of this paper had any conflict of interest that could 430 inappropriately influence the presented study. 431 432 Acknowledgment 433 434 The authors would like to thank the DPI2013-44371-P and the DPI2016-80796-435 9 Projects at the Spanish Ministry of Economics and Competitiveness; and the 436 Telefonica Chair “Intelligence in Networks” of the University of Seville, Spain; for 437 funding our study. 438 439 Funding 440 441 This study was funded by the DPI2013-44371-P and the DPI2016-80796-9 442 Projects at the Spanish Ministry of Economics and Competitiveness; and the Telefonica 443 Chair “Intelligence in Networks” of the University of Seville, Spain. 444 Journal of Biomechanical Engineering 32 Corresponding author: Carmen Ridao-Fernández [36] Roijezon, U., Clark, N. C., and Treleaven, J., 2015, “Proprioception in 558 musculoskeletal rehabilitation. Part 1: Basic science and principles of assessment 559 and clinical interventions.,” Man. Ther., 20(3), pp. 368–377. 560 [37] Dedieu, P., and Zanone, P. G., 2012, “Effects of gait pattern and arm swing on 561 intergirdle coordination.,” Hum. Mov. Sci., 31(3), pp. 660–671. 562 [38] O’Connor , C. M., Thorpe, S. K., O’Malley, M. J., Vaughan, C. L., 2007, 563 “Automatic detection of gait events using kinematic data,” Gait Posture, vol. 25, 564 pp. 469–474. 565 Journal of Biomechanical Engineering 33 Corresponding author: Carmen Ridao-Fernández Supplier 566 567 SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606. 568 Journal of Biomechanical Engineering 34 Corresponding author: Carmen Ridao-Fernández Figure Captions List 569 570 Fig. 1 Assisted gait was contralateral with simultaneous support for heel and crutch at 571 two points 572 Fig. 2 GCH 2.0 Load Measurement System for assisted gait with forearm crutches 573 574 Fig. 3 Representation of the parameters: step length, step angle and step width 575 Stride length was defined as the distance between one heel strike and the next strike 576 with the same heel. The heel strike event was determined analyzing the vertical 577 component of the foot velocity and the trajectory of the marker placed on the heel 578 following the foot velocity algorithm (FVA) proposed by O’Connor [38]. This 579 procedure employed the markers HEE and TOE placed on the calcaneus and over the 580 second metatarsal head, respectively. Stride period was defined as the time to carry out 581 one stride. Stride velocity was defined as the ration between the stride length and the 582 stride period. Step length was defined as the length between the heel strike of one foot 583 and the next heel strike carried out by the other foot. Step period was defined as the time 584 to carry out one step. Step angle was defined by the anterior-posterior axis of the foot 585 local frame and the line of progression. Anterior-posterior axis was determined by 586 markers HEE and TOE. The step width was defined as the mediolateral distance 587 between the feet and measured on the heel markers. 588 589 Fig. 4 Descriptive analysis representation 590 Journal of Biomechanical Engineering 35 Corresponding author: Carmen Ridao-Fernández Table Captions List 591 592 Table 1 GCH 2.0 Load Measurement System: percentages of body weight applied to 593 the crutch during UAG 594 595 Table 2 GWC, gait without crutches; UAG, unilateral assisted gait in which a 596 comfortable load is applied; Dominant crutch, crutch carried out with the dominant 597 hand; Non-dominant crutch, crutch carried out with the non-dominant hand; CI, 598 confidence interval; , Significant at a 99% confidence level; , Significant at a 95% 599 confidence level; Ns, not significant. 600 601 Table 3 GWC, gait without crutches; UAG, unilateral assisted gait in which a 602 comfortable load is applied; DC, crutch carried out with the dominant hand; NDC, 603 crutch carried out with the non-dominant hand; CI, confidence interval; , Significant 604 at a 99% confidence level; , Significant at a 95% confidence level; Ns, not 605 significant. 606 607 Table 4 GWC, gait without crutches; UAG, unilateral assisted gait in which a 608 comfortable load is applied; DC, crutch carried out with the dominant hand; NDC, 609 crutch carried out with the non-dominant hand; CI, confidence interval; , Significant 610 Journal of Biomechanical Engineering 36 Corresponding author: Carmen Ridao-Fernández at a 99% confidence level; , Significant at a 95% confidence level; Ns, not 611 significant. 612 613 Table 5 GWC, gait without crutches; UAG, unilateral assisted gait in which a 614 comfortable load is applied; DC, crutch carried out with the dominant hand; NDC, 615 crutch carried out with the non-dominant hand; CI, confidence interval; , Significant 616 at a 99% confidence level; , Significant at a 95% confidence level; Ns, not 617 significant. 618 619 Table 6 GWC, gait without crutches; UAG, unilateral assisted gait in which a 620 comfortable load is applied; DC, crutch carried out with the dominant hand; NDC, 621 crutch carried out with the non-dominant hand; CI, confidence interval; , Significant 622 at a 99% confidence level; , Significant at a 95% confidence level; Ns, not 623 significant. 624 625 Table 7 GWC, gait without crutches; UAG, unilateral assisted gait in which a 626 comfortable load is applied; DC, crutch carried out with the dominant hand; NDC, 627 crutch carried out with the non-dominant hand; CI, confidence interval; , Significant 628 at a 99% confidence level; , Significant at a 95% confidence level; Ns, not 629 significant. 630 Journal of Biomechanical Engineering 37 Corresponding author: Carmen Ridao-Fernández 631 Fig. 1 Assisted gait was contralateral with simultaneous support for heel and 632 crutch at two points 633 Journal of Biomechanical Engineering 38 Corresponding author: Carmen Ridao-Fernández 634 Fig. 2 GCH 2.0 Load Measurement System for assisted gait with forearm crutches 635 Journal of Biomechanical Engineering 39 Corresponding author: Carmen Ridao-Fernández 636 Fig. 3 Representation of the parameters: step length, step angle and step width 637 Stride length was defined as the distance between one heel strike and the next strike 638 with the same heel. The heel strike event was determined analyzing the vertical 639 component of the foot velocity and the trajectory of the marker placed on the heel 640 following the foot velocity algorithm (FVA) proposed by O’Connor [38]. This 641 procedure employed the markers HEE and TOE placed on the calcaneus and over the 642 second metatarsal head, respectively. Stride period was defined as the time to carry out 643 one stride. Stride velocity was defined as the ration between the stride length and the 644 stride period. Step length was defined as the length between the heel strike of one foot 645 and the next heel strike carried out by the other foot. Step period was defined as the time 646 to carry out one step. Step angle was defined by the anterior-posterior axis of the foot 647 local frame and the line of progression. Anterior-posterior axis was determined by 648 markers HEE and TOE. The step width was defined as the mediolateral distance 649 between the feet and measured on the heel markers. 650 Journal of Biomechanical Engineering 40 Corresponding author: Carmen Ridao-Fernández 651 Fig. 4 Descriptive analysis representation 652 Journal of Biomechanical Engineering 41 Corresponding author: Carmen Ridao-Fernández Table 1 GCH 2.0 Load Measurement System: percentages of body weight applied to 653 the crutch during UAG 654 Percentages of body weight applied to the crutch Subject Mean (%) Min/Max (%) 1 9.48 8.23/11.78 2 9.87 9.23/12.26 3 10.19 7.07/11.59 4 10.91 9.12/11.04 5 9.76 8.82/10.91 6 10.49 9.64/10.98 7 9. . 32 9.12/10.86 655 Journal of Biomechanical Engineering 48 Corresponding author: Carmen Ridao-Fernández the non-dominant hand; CI, confidence interval; , Significant at a 99% confidence 676 level; , Significant at a 95% confidence level; Ns, not significant. 677 Journal of Biomechanical Engineering 49 Corresponding author: Carmen Ridao-Fernández Table 6 Step angle analysis 678 STEP ANGLE Descriptive analysis of step angle (degrees) Right step angle Left step angle Subject Mean (SD) Min/Max Percentiles Mean (SD) Min/Max Percentiles 25 50 75 25 50 75 1 GWC 29.99(2.77) 26.02/35.17 27.82 30.13 31.28 GWC 22.57(3.86) 16.48/29.75 20.00 22.61 25.20 DC 30.66(2.03) 25.63/33.88 30.66 30.66 31.53 DC 31.64(0.75) 29.94/33.00 31.59 31.64 31.78 NDC 33.74(0.00) 33.74/33.74 33.74 33.74 33.74 NDC 27.36(0.00) 27.36/27.36 27.36 27.36 27.36 2 GWC 21.04(4.01) 10.88/25.54 20.43 21.99 23.12 GWC 15.43(1.30) 13.16/17.99 14.54 15.51 15.92 DC 18.99(1.20) 16.40/20.22 18.54 18.99 20.06 DC 20.30(1.25) 18.15/22.69 19.51 20.30 20.83 NDC 24.11(0.53) 22.99/25.24 24.11 24.11 24.11 NDC 16.49(0.94) 14.50/18.47 16.49 16.49 16.49 3 GWC 30.66(2.70) 26.13/35.62 28.72 31.10 32.16 GWC 27.12(4.59) 21.58/34.95 22.82 26.36 31.11 DC 18.19(2.63) 13.62/21.47 15.66 18.27 20.85 DC 23.82(2.54) 18.55/27.58 22.39 24.42 25.45 NDC 26.22(1.41) 23.23/27.58 25.15 26.56 27.36 NDC 21.84(3.92) 16.59/27.85 18.43 21.39 24.97 4 GWC 32.12(3.62) 24.20/37.34 31.11 31.47 35.07 GWC 19.06(3.88) 15.38/29.40 17.15 18.01 19.54 DC 19.64(2.98) 15.05/24.99 17.48 19.32 22.40 DC 25.59(2.30) 20.84/28.21 24.30 26.20 27.51 NDC 26.22(1.42) 23.74/27.64 24.81 26.72 27.29 NDC 18.59(2.48) 15.45/23.75 16.23 18.68 19.87 5 GWC 28.55(3.11) 22.66/31.97 25.69 29.10 31.32 GWC 25.10(3.50) 19.49/31.45 22.31 25.10 27.31 DC 26.65(5.79) 16.58/33.45 24.36 26.65 32.05 DC 24.41(1.34) 22.50/26.62 23.42 24.41 25.00 NDC 28.24(1.19) 26.52/30.01 27.19 28.24 29.20 NDC 19.72(1.38) 17.90/22.54 18.31 19.72 20.65 6 GWC 25.13(2.68) 21.51/30.63 23.30 25.13 26.03 GWC 16.60(3.82) 12.42/26.14 13.54 16.51 17.04 DC 18.40(6.26) 10.42/29.32 13.91 16.51 24.17 DC 23.85(2.28) 20.18/27.99 22.30 23.71 25.18 NDC 27.11(2.34) 21.28/30.51 27.11 27.11 28.05 NDC 24.90(2.93) 17.76/28.89 24.89 24.90 25.69 7 GWC 21.95(0.00) 21.95/21.95 21.95 21.95 21.95 GWC 10.44(0.00) 10.44/10.44 10.44 10.44 10.44 DC 20.32(3.44) 15.20/28.23 18.60 20.32 21.52 DC 19.86(2.45) 13.79/22.42 19.44 19.86 21.72 NDC 25.18(0.23) 24.69/25.66 25.18 25.18 25.18 NDC 10.85(0.32) 10.18/11.53 10.85 10.85 10.85 Step angle analysis. Difference of means between GWC and UAG with DC or NDC Right step angle Left step angle Subject DC DC CI of the difference of means (m) p Effect size CI of the difference of means (m) p Effect size 1 - 3.396;2.055 Ns - 12.095; - 6.055 0.005 0.887 2 - 0.162;4.267 Ns - 5.759; - 3.983 <0.001 0.886 3 9.342;15.614 <0.001 0.920 0.185;6.408 0.040 0.406 4 8.390;16.587 <0.001 0.883 - 10.630; - 2.421 0.017 0.757 5 - 3.403;7.194 Ns - 2.001;3.390 Ns 6 1.893;11.570 0.012 0.573 - 10.329; - 4.192 0.007 0.854 7 - - - - Subject NDC NDC CI of the difference of means (m) p Effect size CI of the difference of means (m) p Effect size 1 - - - - 2 - 5.987; - 0.153 0.022 0.726 - 2.220;0.099 Ns 3 2.509;6.373 0.001 0.718 1.655;8.893 0.009 0.525 4 2.720;9.091 0.007 0.854 - 2.930;3.869 Ns 5 - 1.768;2.386 Ns 3.079;7.682 0.001 0.711 6 - 5.294;1.328 Ns - 13.006; - 3.597 0.021 0.727 7 - - - - GWC, gait without crutches; UAG, unilateral assisted gait in which a comfortable load 679 is applied; DC, crutch carried out with the dominant hand; NDC, crutch carried out with 680 Journal of Biomechanical Engineering 50 Corresponding author: Carmen Ridao-Fernández the non-dominant hand; CI, confidence interval; , Significant at a 99% confidence 681 level; , Significant at a 95% confidence level; Ns, not significant. 682 Journal of Biomechanical Engineering 51 Corresponding author: Carmen Ridao-Fernández Table 7 Step width analysis 683 STEP WIDTH Descriptive analysis of step width (m) Subject Mean (SD) Min/Max Percentiles 25 50 75 1 GWC 0.10(0.10) 0.02/0.35 0.03 0.09 0.13 DC 0.04(0.01) 0.03/0.05 0.04 0.04 0.04 NDC 0.07(0.00) 0.07/0.07 0.07 0.07 0.07 2 GWC 0.06(0.02) 0.04/0.11 0.05 0.05 0.06 DC 0.06(0.01) 0.03/0.07 0.05 0.06 0.06 NDC 0.06(0.00) 0.06/0.06 0.06 0.06 0.06 3 GWC 0.11(0.03) 0.08/0.16 0.09 0.10 0.15 DC 0.12(0.02) 0.08/0.14 0.11 0.12 0.13 NDC 0.12(0.02) 0.09/0.15 0.10 0.13 0.14 4 GWC 0.08(0.01) 0.06/0.10 0.07 0.08 0.09 DC 0.10(0.02) 0.07/0.12 0.08 0.10 0.11 NDC 0.18(0.12) 0.11/0.53 0.13 0.14 0.16 5 GWC 0.08(0.03) 0.04/0.14 0.07 0.08 0.10 DC 0.05(0.01) 0.04/0.08 0.04 0.05 0.05 NDC 0.06(0.01) 0.04/0.09 0.06 0.06 0.07 6 GWC 0.09(0.02) 0.06/0.13 0.08 0.09 0.11 DC 0.09(0.02) 0.07/0.12 0.07 0.07 0.10 NDC 0.09(0.01) 0.08/0.10 0.09 0.09 0.09 7 GWC 0.07(0.00) 0.07/0.07 0.07 0.07 0.07 DC 0.06(0.03) 0.00/0.11 0.04 0.06 0.09 NDC 0.07(0.00) 0.06/0.07 0.07 0.07 0.07 Step width analysis. Difference of means between GWC and UAG with DC or NDC Subject DC CI of the difference of means (m/s) P Effect size 1 - 0.004;0.135 0.028 0.693 2 - 0.012;0.020 Ns 3 - 0.030;0.020 Ns 4 - 0.031;0.001 Ns 5 0.016;0.054 0.009 0.823 6 - 0.012;0.026 Ns 7 - - Subject NDC CI of the difference of means (m/s) P Effect size 1 - - 2 - 0.018;0.010 Ns 3 - 0.037;0.020 Ns 4 - 0.185; - 0.006 0.005 0.886 5 - 0.003;0.043 Ns 6 - 0.013;0.016 Ns 7 - - GWC, gait without crutches; UAG, unilateral assisted gait in which a comfortable load 684 is applied; DC, crutch carried out with the dominant hand; NDC, crutch carried out with 685 Journal of Biomechanical Engineering 52 Corresponding author: Carmen Ridao-Fernández the non-dominant hand; CI, confidence interval; , Significant at a 99% confidence 686 level; , Significant at a 95% confidence level; Ns, not significant. 687