Three-dimensional motion analysis of the trailing limb in obstacle crossing during gait in young adults: foot trajectory changes with the height of obstacles

Abstract

[Purpose] We aimed to analyze the kinematic characteristics of the foot trajectory of the trailing limb in a crossing motion during gait when the height of an obstacle is varied. [Participants and Methods] The participants were 12 healthy adult males (age: 24.2 ± 0.9 years). The participants performed 11 tasks: normal gait and crossing an obstacle with a height ranging from 10% to 100% of the height from the floor to the knee joint crease at 10% intervals during gait. The parameters included the clearance height, foot trajectory, swing phase duration, stride, toe distance, and heel distance. [Results] In the crossing condition, the foot trajectory changed with the obstacle height, with some participants showing unimodal trajectories when the obstacle height was ≥30% of the knee joint height. The trend test results show a trend from bimodal to unimodal trajectories with increasing obstacle heights. [Conclusion] The trailing limb of a young participant’s crossing motion during gait showed kinematic changes in the foot trajectory, adjusting from bimodal to unimodal as the obstacle height increased, suggesting that this may be a specific strategy for dealing with high obstacles.

INTRODUCTION

Falls in the elderly are often caused by not only age-related decline in motor function but also environmental factors¹⁾. Daily activities are predictive behaviors. Falls occur when the environment is contrary to expectations when attempting a task, when physical movements cannot be performed as predicted, or when the environment is not well understood and appropriate predictions cannot be made²⁾. A tendency to avoid rather than cross when passing over them has been reported as the higher the obstacle³⁾, but stumbles and falls are thought to occur when errors in environmental constraints, anticipatory behavior, or understanding of the environment happen in this method of avoidance. Among obstacle passing movements, many researchers are working to elucidate the characteristics of the crossing motion that causes stumbling. In the clinical setting of physical therapy, patients crossing motion are commonly the target of interventions and evaluations to improve their functional gait ability and adaptation to daily living. However, unlike walking with plain movements, crossing movements are influenced by environmental factors and are complex in terms of physical control. Especially when crossing an obstacle on a trailing limb, the leg and obstacle are out of sight. Therefore, clearance of obstacles fluctuates in the elderly, and the difficulty is considered high⁴⁾. In the study by Heijnen et al.⁵⁾ in young adults, a hypothetical crossing movement with an obstacle after performing and memorizing the crossing movement 25 times, and found that contact with the hypothetical obstacle was 9% in the leading limb and 47% in the trailing limb. This report suggested that obstacle recognition and memory during obstacle approach are important in the trailing limb. They also had 15 young participants perform an obstacle crossing task while gait 300 trials in a row, reported that 12 of the 15 participants stumbled at least once, and 22 of the 24 stumbles were caused by trailing limbs, with an average decrease in trailing limb clearance of 1 mm per trial until the stumble occurred⁶⁾. Focusing on the trailing limb is clinically important, Hsu et al. analyzed the crossing motion of middle-aged people with type 2 diabetes who had no or mild peripheral neuropathy, and reported that the clearance height of the trailing limb was significantly lower than that of healthy middle-aged people, accompanied by changes in the angle of the hip and knee joints, therefore increasing the risk of stumbling⁷⁾. The role of traversing the trailing limb during gait is not only to ensure safety to cross obstacles, but also to smoothly return to gait after crossing an obstacle, requiring an efficient movement strategy. We considered that lowering the clearance would certainly lead to a reduction in energy costs, but that it could also have the potential to lead directly to a failure in ensuring safety, such as stumbling over an obstacle. Therefore, we thought it would be useful to examine whether there is any change in the method of operation after safety is ensured. In the hypothesis of energy cost adjustment, there is a possibility that some other kinematic characteristics may be observed in the movement of the trailing limb, which is the stage of transition from crossing motion to gait, and that analysis and clarification of these characteristics would be relevant to clinical assessment and intervention of crossing motion, making the study highly significant. The purpose of this study was to focus on the kinematic changes of the trailing limbs in the crossing motion when the height of the obstacle is changed, and to analyze the foot trajectory that is considered to be meaningful in general use, especially because it is likely to be visually observed without using special equipment in a clinical setting.

PARTICIPANTS AND METHODS

Twelve healthy young adults with no history of orthopedic or neurological diseases were included in this study. They were 24.2 ± 0.9 years old, 174.3 ± 5.6 cm tall, 65.1 ± 6.0 kg in weight, and 46.9 ± 2.5 cm in knee height.

Participants were recruited at the collaborator’s facility by posting participant recruitment guidelines. The purpose and methods of the study were explained to them in detail. Participants were asked to undergo measurements if they agreed to participate in the study. The purpose and methods of the study were explained to all the participants in accordance with the Declaration of Helsinki, and verbal and written informed consent was obtained. This study was approved by the Ethical Review Committee of Hidaka Rehabilitation Hospital (Ethical Review Committee No. 161101 of Hidaka Rehabilitation Hospital).

A three-dimensional motion analyzer (Locus 3D, ANIMA, Tokyo, Japan, 10 infrared cameras) and a sheet-type lower limb load cell (Walk Way MW-1000, ANIMA) were synchronized as measurement devices. Infrared reflective markers were attached to the heel (lateral process of the calcaneal ridge) and toe (nail of the thumb) on both the left and right sides. The sampling frequency was 100 Hz, and the measurement time was 10 s. The data obtained from the three-dimensional motion analysis analyzer were low-pass filtered at 10 Hz. The walking path was 10 m long and 2 m wide with an obstacle placed in the middle. The obstacle was a flat aluminum bar, 1 cm deep and 1 mm thick, placed on two columns on either side. The heights of the bars were adjusted arbitrarily.

The measurements included the trailing limb clearance height at the time of crossing an obstacle during gait, swing phase duration, stride, toe distance (TD), heel distance (HD), and foot trajectory. The clearance height in the gait condition was the minimum value on the sagittal plane of the height of the toe marker during the swinging leg, whereas in the crossing condition, the height was the distance between the marker and the obstacle (vertical distance from the obstacle) at the time the marker was positioned directly over the obstacle (Fig. 1) and was normalized to body height. The swing phase duration was defined as the time from when the trailing limb left the ground before crossing the obstacle to when it touched the ground after crossing the obstacle. It was not specified which foot position was used for ground release or ground contact. Foot release and grounding were determined based on the position of the heel and toe markers on the vertical axis in the resting standing position. Grounding was defined as the last time either marker was greater than their positions before the obstacle point was crossed, and grounding was defined as the first time either marker was less than their positions after the obstacle point was crossed. The stride was defined as the distance between the heels of the trailing limbs before and after crossing. TD was defined as the distance from the toe position at the time of toe-off before crossing over to the obstacle. HD was defined as the distance from the obstacle to the heel of the trailing limb after crossing over (Fig. 1). The foot trajectory was defined as a unimodal trajectory when the sagittal trajectory of the toe marker was plotted with one upward convex peak value (local maximum) in one location and bimodal trajectory when it was plotted in two locations. Examples of the unimodal and bimodal trajectories are shown in Fig. 2.

Fig. 1.

Clearance height, toe distance (TD), heel distance (HD) and stride.

Fig. 2.

Foot trajectory.

The solid line shows a bimodal trajectory with two maxima (▲) and a minimum (▽), and the dotted line shows a unimodal trajectory with only one maximum.

Before to the measurement, the participants were allowed to walk comfortably an arbitrary number of times without obstacles and understand and adjust the position of the obstacles during the first few steps of walking. The participants were instructed to cross over the obstacle as they normally do in daily life and were told, “Please walk at your usual comfortable walking speed and cross over the obstacle as you normally do”. After crossing over an obstacle, continue to walk to the end of the walkway. However, in this study, the right leg was used as the leading limb to cross the obstacle (or a line drawn on the floor in the gait condition).

The height of the obstacle was set in 10 conditions with 10% intervals between 10–100% of the height from the ground to the lateral knee joint crease in the standing position. Measurements were obtained once during normal gait without obstacles (gait condition), followed by randomization to the condition in which the participant crossed an obstacle while gait (crossing condition). To eliminate learning effects and fatigue, we recorded single trial for each condition (11 trials in total). No practice trials were conducted before the experimental sessions. The order of the crossing conditions was randomized between the participants using a random number table.

For statistical analysis, each measurement in the 11 conditions was tested by repeated measures analysis of variance using SPSS statistics22.0. Tukey’s multiple comparison test was performed as a post hoc test. To analyze the trend of foot trajectories, Fisher’s exact test and the Cochran–Armitage trend test were performed using the test software R2.8.1. The significance level was set at 5%.

RESULTS

Table 1 presents the clearance height, swing phase duration, stride, TD, HD, and foot trajectory results. Analysis of variance showed a significant difference in clearance height (F=10.5, p<0.05) between the gait and crossing conditions; however, there was no significant difference within the crossing conditions. The analysis of variance showed a significant difference in the swing phase duration (F=57.3, p<0.05) between the gait and crossing conditions, and a significant difference from the gait condition was found in the crossing condition with an obstacle height of 20% or higher. Significant differences in swing phase duration were observed in the crossing condition for obstacle heights of 10% and 40% or more; 20% and 50% or more; 30% and 60% or more; 40% and 10% or 80% or more; 50% and 10% or 80% or more; 60% and 30% or less or 80% or more; 70% and 30% or less or 90% or more obstacle height, 80% and 60% or less or 100%; 90% and 70% or less; and 100% and 80% or less. The analysis of variance was not significant for stride, TD, or HD. Under gait conditions, the foot trajectory was bimodal in all cases. In the crossing condition, the foot trajectory changed with the obstacle height, with some participants showing unimodal trajectories when the obstacle height was 30% or greater, and a trend from bimodal to unimodal trajectories was observed with increasing obstacle height (χ²=30.0, p<0.05).

Table 1. Clearance height, swing phase duration, stride, TD, HD and foot trajectory.

Conditions		Clearance height	Swing phase duration	Stride (cm)	TD (cm)	HD (cm)	Foot trajectory (person)
		(%)	(seconds)				Bimodal	Unimodal
Gait		2.9 ± 0.6b–k	0.39 ± 0.02c–k	138.7 ± 8.8	19.6 ± 6.4	92.8 ± 11.7	12	0
Crossing	10%	10.0 ± 3.2a	0.45 ± 0.06e–k	136.6 ± 7.3	20.3 ± 7.0	90.0 ± 10.7	12	0
(Obstacle	20%	10.2 ± 3.4a	0.47 ± 0.05a,f–k	138.4 ± 9.7	22.5 ± 6.6	89.3 ± 8.6	12	0
height)	30%	9.9 ± 3.3a	0.50 ± 0.07a,g–k	138.8 ± 6.6	20.3 ± 5.3	92.3 ± 9.8	10	2
	40%	10.0 ± 3.3a	0.53 ± 0.07a,b,i–k	141.0 ± 7.7	21.7 ± 6.3	93.1 ± 10.6	11	1
	50%	10.7 ± 3.4a	0.53 ± 0.06a–c,i–k	139.7 ± 6.2	22.0 ± 7.5	91.6 ± 11.2	9	3
	60%	10.7 ± 4.3a	0.57 ± 0.09a–d,i–k	138.9 ± 5.8	23.8 ± 8.0	88.9 ± 9.8	8	4
	70%	11.2 ± 4.4a	0.59 ± 0.09a–d,j,k	140.9 ± 7.4	22.7 ± 6.8	91.8 ± 7.8	7	5
	80%	11.1 ± 3.5a	0.65 ± 0.12a–g,k	138.2 ± 8.4	21.4 ± 6.8	90.8 ± 9.3	7	5
	90%	10.4 ± 4.3a	0.67 ± 0.12a–h	138.9 ± 8.0	21.3 ± 5.5	91.5 ± 6.9	5	7
	100%	11.1 ± 4.4a	0.73 ± 0.10a–i	143.5 ± 10.3	22.1 ± 6.8	95.5 ± 8.6	3	9

Mean ± standard deviation.

In Tukey’s multiple comparisons, significantly different from a: gait condition, b: 10%, c: 20%, d: 30%, e: 40%, f: 50%, g: 60%, h:70%, i: 80%, j: 90%, k: 100%.

TD: toe distance; HD: heel distance.

DISCUSSION

The results for measurements other than foot trajectory were generally like those of previous studies. The results suggest that the measurement methods used in the present study are like those used in previous studies of healthy young adults who cross visible obstacles at a sufficient distance.

The clearance height of the trailing limb in the crossing condition was higher than that in the gait condition, but the results showed that the clearance height was constant regardless of obstacle height. Similar results have been reported in the literature^8,9,10) for the leading limb, and in a study on the trailing limb in young adults, Chou and Draganich¹¹⁾ investigated the clearance heights of the gait and crossing conditions at obstacle heights ranging from 5.1 cm to 20.4 cm, and found that the clearance heights were 3.1 cm in the gait condition and 14.6 cm in the crossing condition, and no difference in clearance height due to obstacle height was observed. The results of the present study support those of Chou and Draganich.

The swing phase duration was not significantly different between the near conditions but was significantly different between 20% and 40% of the conditions. The mean tended to increase from the gait condition to the high obstacle height crossing condition, suggesting that the swing phase duration time increased with increasing obstacle height. Ae et al.⁸⁾ reported that the swing phase duration increased in response to an increase in obstacle height in the leading limb in both groups without significant difference between the youth and elderly groups. Similar results were obtained in the trailing limb of healthy young adults as in the leading limb.

No differences in stride, TD, or HD were observed under any condition. Similar results were obtained in previous studies^{12, 13)}, which also found no significant differences in the TD and HD as a function of the obstacle height. In addition, analyses by Hakamata et al.^{14, 15)} and Chou and Draganich¹⁶⁾ on TD and HD have reported that interventions that bring the leading limb foot closer to the obstacle after crossing may improve obstacle avoidance by the trailing limb in young adults and healthy elderly individuals. In these studies, clearance and collision rate with obstacles were measured in such a way as to consciously adjust the TD or HD position, but in the present study, the foot-ground position was not specified. In addition, practice was provided in advance to avoid the need for preparatory strategies for obstacle avoidance by adjusting stride length depending on the obstacle position. It is possible that the trial did not have enough difficulty to make the participants change their foot-ground position to avoid the obstacle.

To date, the measurement results have generally been similar to the mainstream results of previous studies and research on leading limbs. Tokuda and Saikawa¹⁷⁾ stated that crossing differs from gait because safety is a more important constraint than energy cost optimization. In the present study, it was evident that the participants performed the movement by raising the lower limb over time rather than adjusting the stride length to maintain a certain clearance height to ensure the safety of the crossing movement. Moraes et al.¹⁸⁾ reported that when there is sufficient time after obstacle detection, stability is less affected because a rapid change in stride length is not required, and forward movement is preferred in young adults. Heijnen et al.⁶⁾ reported that in a task in which participants crossed the same obstacle height 300 times, the clearance of the trailing limb progressively decreased until contact with the obstacle was made to minimize the energy cost. Therefore, it is reasonable to assume that young adults’ crossing movements do not arbitrarily secure the swing phase duration for the sake of safety, but extend the swing phase duration to secure the optimal amount of clearance to the extent that the energy cost is not excessive while minimizing the loss of forward motion.

Thus far, the results are generally similar to those of previous studies; however, the results show new findings regarding foot trajectories. In the gait condition, the foot trajectory was bimodal, whereas in the crossing condition, the trajectory changed from bimodal to unimodal as the obstacle height increased above 30% of the obstacle height. Foot trajectory was unimodal when the participants moved from a stationary standing position to a stationary standing position without walking. Unimodal adjustment of the foot trajectory in response to an increase in obstacle height during gait may be a strategy specific to crossing movements to cope with high obstacles. Kunimune and Okada¹⁰⁾ measured 14 young and 14 elderly participants, and the results were different from the present study in that the trailing limb clearance height of the young participants was lower than in the present study and decreased with the height of the obstacle. However, interesting results were obtained, and they reported that in the approaching obstacle, leading limb crossing section, and trailing limb crossing section, postural instability to the side was greatest in the trailing limb crossing section, and the higher the obstacle height, the more pronounced the instability. Yamagata et al.¹⁹⁾ reported that the conservative strategy of raising the trailing limb is not necessarily useful for successful obstacle crossing movements in elderly participants because the inter-articular coordination for controlling the center of gravity position is weaker in the trailing limb than in the leading limb and becomes weaker as the clearance height of the trailing limb increases. The difference in bimodal or unimodal foot trajectories between participants, even at the same obstacle height, may be due to the ability of dynamic postural control to maintain stability. In addition, it was found that a certain amount of clearance from the obstacle was secured in the trailing limb of the cross-movement, but this required the same degree of lower-limb elevation as the obstacle height increased. Therefore, muscle activity, balance function, and flexibility are required to control the vertical transfer of lower limb weight according to the obstacle height on both the stance and swing sides. Differences in physical function are often observed in the elderly, depending on whether they stumble or fall^{20, 21)} and it is undeniable that differences in dynamic postural control ability can cause kinematic changes even in young people, even if they do not fall. Furthermore, Patla and Rietdyk⁹⁾ performed an obstacle crossing tasks while gait on 6 young participants at different heights (6.7 cm, 13.4 cm, and 26.8 cm) and reported that the risk of tripping or slipping was minimized by decreasing the horizontal velocity of the leading foot and increasing the vertical velocity in response to an increase in obstacle height when the leading limb contacted the ground. If a similar strategy is chosen for the trailing limb to ensure safety during ground contact, the foot trajectory will likely become unimodal, as the rate of descent from the maximum foot height increases as the limb passes over the obstacle.

Based on the results of this study, if the trajectory of the trailing foot of the crossing during gait is bimodal, it is most likely that the participant has adopted a strategy that prioritizes a smooth return to gait because safety and stability are under control. Conversely, if it is unimodal, it is likely that the patient is making kinematic changes specific to crossing to control safety and stability, rather than forward motion. Although additional research is needed, this suggests that this may be a useful point of reference for setting the level of difficulty when performing crossing motions during gait in physical therapy interventions. For example, by focusing on foot trajectory when practicing gait exercises including crossing motion, it may be possible to evaluate whether the height of the obstacle was optimal for setting the difficulty level, what the difficulty level of the obstacle is for the participant, and the priority level of efficiency of the participant’s movement (When conducting functional gait exercises for elderly people who have difficulty recognizing obstacles, it is considered problematic that they tend to walk in a bimodal pattern even when the obstacle conditions are high, etc.). In addition, it may be possible to reduce the risk of falls by teaching modifications to the method of movement. Conversely, in cases where physical function is high but there is a strong fear of motion or a delay in improving motion ability, dynamic balance exercise to promote smooth motion execution may lead to reduced energy costs by using instruction and exercise that results in a bimodal trajectory.

The limitations of this study include the fact that only the distance and trajectory between the toe and heel markers and the obstacle were shown and could not relate to control at each joint, and the fact that it may not be directly applicable to elderly participants because the participants were able-bodied. In the bimodal trajectory during gait, heel-ground contact in the dorsiflexed ankle position at the beginning of stance is considered to represent the second local maximum value of the foot apex. However, it remains to be examined whether the bimodal trajectory during crossing an obstacle is due to heel-ground contact in the dorsiflexed ankle position as in walking. To clarify the functional significance of the foot trajectory, joint alignment analysis, electromyography, and electromyography should be conducted. To clarify the functional significance of foot trajectory, verification from the aspects of joint alignment analysis, electromyography, floor reaction force analysis, etc. will be required in the future. Further clarification of this issue is expected to lead to further verification of at what point in the crossing motion during gait due to the height of the obstacle, safety is prioritized over energy cost minimization. In addition, the results of the post-hoc analysis of 1-β power showed that the analysis of variance was inadequate for some measurements, and that the sample size was problematic due to inadequate power in multiple comparisons. Statistical validation should be deepened with a sample size that has sufficient test power. In foot trajectory, it is necessary to clarify whether the trend implied by the difference can be indicated by a cutoff, and whether it is a stepwise or linear response. In the future, a more detailed kinematic analysis, including joint motion, kinematic analysis using a ground reaction force meter, and a comparison between the elderly and persons with disabilities, is needed.

Conference presentation

Parts of this paper were presented at the 37th Annual Conference of Physical Therapists of Kanto Koshinetsu Block and used in the abstract of the same conference (https://doi.org/10.14901/ptkanbloc.37.0_0048).

Conflicts of interest

There are no conflicts of interest to disclose regarding this paper.