Effect of experimentally induced plantar pain on trunk posture during gait

Abstract

[Purpose] Plantar pain is associated with the prevalence of low back pain. Therefore, it is reasonable to assume that some kind of physical change should be occurring in the trunk due to plantar pain. However, the physical effect of plantar pain on the trunk remains unknown. We evaluated the effect of plantar pain on trunk posture during gait. [Participants and Methods] Ten healthy volunteers participated in the present study. Participants walked under two conditions: without pain and with pain. In the with pain condition, we set pain-inducing devices to the right foot to induce plantar pain during stance phase. By using 3D motion analysis system, the angles of the head, thorax, and pelvis segments, as well as the neck, trunk, bilateral hip, bilateral knee, and bilateral ankle joints, were measured. We analyzed the angle data throughout the gait cycle by using one-dimensional statistical parametric mapping. [Results] The anterior trunk tilt was observed in the right stance phase. [Conclusion] The anterior trunk tilt observed in the with pain condition may be a burden on the trunk. Our results presented one of the possible reasons for increased prevalence of low back pain in the plantar pain patients.

INTRODUCTION

Foot pain has been reported as widespread in the general population, affecting approximately 20% of adults^1,2,3). For example, plantar fasciitis is one of the major reasons for foot plantar pain. Excessive running, prolonged standing, and obesity are risk factors for developing this condition^{4, 5)}. The discomfort of foot plantar pain may not only be limited to the foot but may also cause pain in other parts of the body. In fact, it has been shown that foot plantar pain increases the prevalence of low back pain⁶⁾.

Because foot plantar pain increases the prevalence of low back pain, it is reasonable to assume that some kind of physical change should be occurring in the trunk due to plantar pain. However, the physical effect of plantar pain on the trunk remains unknown. Previous studies have revealed that plantar pain causes changes in the joint angles in the lower limbs during gait: flexion of the painful limb and extension of the painless limb^{7, 8)}. Yet, whether plantar pain causes changes in trunk posture during gait has not been evaluated.

Considering the prevalence of low back pain associated with foot plantar pain⁶⁾, it is important to know whether the trunk posture changes with plantar pain during gait. Therefore, this study aimed to evaluate the effect of plantar pain on trunk posture during gait. We experimentally induced pain in the foot planta. We hypothesized that changes in trunk posture would occur because plantar pain increases the prevalence of low back pain.

PARTICIPANTS AND METHODS

Ten healthy young males (mean [standard deviation]: height 173.3 [10] cm; weight 65.3 [6.4] kg; age 22.8 [1.9] years) participated in the present study. The participants provided written informed consent. Ethical approval was obtained from the Research Ethics Committee, School of Engineering, The University of Tokyo (Approval number: KE21-68).

The participants’ gait was evaluated under two conditions: (a) without pain and (b) with pain in the foot planta. In (a) the without pain condition, the participants walked as usual. In (b) the with pain condition, experimental pain was induced in the right planta. The participants walked at their self-selected speeds, and three trials were performed for each condition. To avoid the possibility of residual pain affecting the gait, (a) gait without pain was performed before (b) gait with pain.

As pain-inducing devices, protrusions with a height of 10 mm were made of acrylonitrile butadiene styrene (ABS) resin using a 3D printer (Stratasys F170; RICOH COMPANY, Ltd., Minato-ku, Tokyo, Japan). For safety reasons, the apexes of the devices were shaved by 2 mm. In (b) with pain condition, three pain-inducing devices were attached to the right shoe, stimulating the thenar eminence, the hypothenar, and the heel. We attached the devices to these locations because we wanted to ensure that pain was induced. If we use fewer devices, the participants may avoid landing on the foot part where the devices are attached. In this case, pain may not be induced. Therefore, we attached three devices. No pain-inducing device was attached to the left shoe. To minimize the influence of the shoes themselves on gait, soft shoes (UX-1031 Marine Shoes; CAPTAINSTAG Co. Ltd., Niigata, Japan) were used for the experiment. In (a) the without pain condition, the participants wore the same shoes without the pain-inducing device to allow a comparison between the two conditions.

We measured i) pain, ii) gait parameters, and iii) segment and joint angles. To measure i) pain, a visual analog scale (VAS; 0: no pain, 100: no more pain) was used to document plantar pain. The VAS score of each participant was measured immediately after each trial. We identified a minimally important difference of VAS as 20⁹⁾. To measure ii) gait parameters, a pressure platform (the FDM-2 System; zebris Medical Gmbh, Isny, Germany) was used. The sampling rate was 100 Hz. The gait cycle, the stance phase, and the swing phase were calculated.

To measure iii) segment and joint angles, a 3D motion analysis system with ten cameras (Mac3D system; Motion Analysis Corporation, Rohnert Park, CA, USA) was used. Infrared reflective markers were attached to the body surface in accordance with the CGM marker set version 2.4¹⁰⁾. The sampling rate was 100 Hz. The segment and joint angles were calculated and low pass filtered (10 Hz, Butterworth filter) using software (Visual 3D; C-Motion, Inc., Germantown, MA, USA, and MATLAB R2022a; The MathWorks Inc., Natick, MA, USA). The angles of head segment, thorax segment, and pelvis segment were calculated in global coordinates (the axis system of the measurement space). The joint angles were calculated as relative angles (in local coordinates) with respect to the segment closer to the pelvis. The calculated joint angles were the neck, trunk, bilateral hip, bilateral knee, and bilateral ankle. Here, the neck angle was calculated as the relative angle between the head segment and the thorax segment. The trunk angle was calculated as the relative angle between the thorax segment and the pelvis segment. The joint flexion/extension angles were analyzed.

For all the data, we first averaged the data from the three trials for each participant and for each condition. The between-participant data were then averaged for each condition. For i) the VAS score data and ii) the gait parameters, the differences in the means between the conditions were analyzed using paired two-tailed t-tests.

For iii) the segment and joint angle data, one-dimensional statistical parametric mapping (1d SPM) was performed. 1d SPM provides a framework for the continuous statistical analysis of a smooth bounded one-dimensional field, using random field theory¹¹⁾. Significant differences in local intervals were calculated when comparing continuous data. The data were analyzed using a paired two-tailed t-test for significant differences between the conditions throughout the gait cycle. A p-value of less than 0.05 was considered statistically significant.

RESULTS

As shown in Table 1, i) the VAS scores in (a) the without pain condition and (b) the with pain condition were 0 and 62.2, respectively. The VAS score significantly increased with pain. None of the participants felt pain in (a) the without pain condition. In (b) the with pain condition, the VAS scores of all participants were higher than 20, i.e., all the participants felt pain.

Table 1. VAS score and gait parameters.

	Without pain		With pain
	Left	Right	Left	Right
			no pain	pain
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
VAS score	0 (0)		62.2# (22.6)
Gait cycle [s]	1.23 (0.10)		1.39 (0.34)
Stance phase [s]	0.80 (0.07)	0.79 (0.08)	1.08# (0.33)	0.89 (0.27)
Stance phase [% GC]	64.8 (1.8)	64.4 (2.3)	77.4# (5.6)	63.6 (4.7)
Swing phase [s]	0.43 (0.04)	0.43 (0.04)	0.30† (0.07)	0.49 (0.08)
Swing phase [% GC]	35.2 (1.8)	35.6 (2.3)	22.6† (5.6)	36.4 (4.7)
Single stance phase [s]	0.44 (0.04)	0.43 (0.44)	0.49 (0.08)	0.30† (0.07)
Single stance phase [% GC]	35.6 (2.3)	35.2 (1.8)	36.5 (5.4)	22.6† (5.8)
Double stance phase [s]	0.36 (0.05)		0.59# (0.27)
Double stance phase [% GC]	29.1 (3.5)		40.8# (8.1)

^#: Significantly larger value in the with pain condition compared to the without pain condition. ^†: Significantly smaller value in the with pain condition compared to the without pain condition. VAS: visual analog scale; GC: gait cycle; SD: standard deviation.

The results of ii) the gait parameters are shown in Table 1 and Fig. 1. In the left lower limb (painless side), the stance phase increased, and the swing phase decreased in (b) the with pain condition. In the right lower limb (painful side in the with pain condition), no significant differences between the conditions were observed in both the stance and swing phases. Although the stance phase of the right lower limb showed no significant difference between the conditions as mentioned above, the single stance phase in the right lower limb decreased with pain. In addition, the double stance phase increased with pain.

Fig. 1.

The stance phase during gait.

The horizontal line shows the time that has elapsed since the left initial contact. The horizontal bold lines indicate the stance phase.The vertical dotted lines indicate the time of the left initial contact. L: left; R: right.

Figure 2 shows the results of iii) the segment and joint angles during the gait cycle in (a) the without pain and (b) the with pain conditions and the results of the 1d SPM analysis. In (b) the with pain condition, anterior tilts in the head, thorax, and pelvis segments were observed in global coordinates during the right stance phase (Fig. 2A–2C. Also, refer to Fig. 1 to see where the right stance phase is in the gait cycle based on the left side). However, the neck and trunk joint angles showed no significant differences between the conditions in local coordinates (Fig. 2D, 2E). Figure 3 shows the results of the joints in the lower limbs. Flexions of the left hip and left knee, as well as plantarflexion of the left ankle, were decreased around the right initial contact (Fig. 3A–3C). Extensions of the right hip and knee, and plantarflexion of the right ankle, were decreased during the right stance phase (Fig. 3D–3F).

Fig. 2.

Mean and standard deviation of the angles and the results of 1d SPM (Head and Trunk).

Upper figure (Angle): The horizontal and vertical lines show the gait cycle and the angle, respectively. The gait cycle was the period from the left initial contact to the next left initial contact. On the vertical lines, a positive value indicates anterior tilt, while a negative value indicates posterior tilt in A, B, and C. Likewise, positive indicates flexion and negative indicates extension in D and E. The line and the upper and lower colored areas show the mean and standard deviation of the data for each condition, respectively. The dotted blue line shows the results under the without pain condition, while the solid red line shows the results under the with pain condition. Lower figure (SPM): The horizontal and vertical lines indicate the gait cycle and t value, respectively. The gait cycle was the period from the left initial contact to the next left initial contact. On the vertical lines, a positive t-value indicates that the with pain condition is more anteriorly tilted/more flexed (or less posteriorly tilted/less extended) than the without pain condition. While, a negative t-value indicates that the without pain condition is more anteriorly tilted/more flexed (or less posteriorly tilted/less extended) than the with pain condition. The red dotted lines represent the significance level (5%). The gray area shows the time series with significant differences between the without pain and with pain conditions. 1d SPM: one dimensional statistical parametric mapping.

Fig. 3.

Mean and standard deviation of the angles and the results of 1d SPM (Lower limb joints).

Upper figure (Angle): The horizontal and vertical lines show the gait cycle and the joint angle, respectively. The gait cycle was the period from the left initial contact to the next left initial contact. On the vertical lines, a positive value indicates flexion, while a negative value indicates extension. The line and the upper and lower colored areas show the mean and standard deviation of the data for each condition, respectively. The dotted blue line shows the results under the without pain condition, while the solid red line shows the results under the with pain condition. Lower figure (SPM): The horizontal and vertical lines indicate the gait cycle and t-value, respectively. The gait cycle was the period from the left initial contact to the next left initial contact. On the vertical lines, a positive t-value indicates that the with pain condition is more flexed (or less extended) than the without pain condition. While, a negative t value indicates that the without pain condition is more flexed (or less extended) than the with pain condition. The red dotted lines represent the significance level (5%). The gray area shows the time series with significant differences between the with pain and without pain conditions. L: left; R: right. d. flexion: dorsiflexion; p. flexion: plantarflexion; 1d SPM: one dimensional statistical parametric mapping.

DISCUSSION

In our present study, healthy participants walked under two conditions: (a) without pain and (b) with pain. In the (b) with pain condition, we experimentally induced pain in the right foot plantar. The i) VAS score, ii) gait parameters, and iii) segment and joint angles were measured. The measurements significantly changed due to pain.

The i) VAS score in (b) the with pain condition was 62.2 (mean). In a previous study that induced experimental pain using an injection of hypertonic saline, the amount of pain reached 6.1 in an 11-point numerical ratio scale (NRS) score¹²⁾. NRS is strongly associated with VAS¹³⁾, and NRS 6.1 in a previous study is near to VAS 62.2 in our present study. We were able to induce almost the same level of pain as in a previous study using a minimally invasive method.

In ii) the gait parameters, the stance phase of the right lower limb (painful side in the with pain condition) showed no significant differences between the conditions. This is compatible with the results of patients with heel pain¹⁴⁾. Although the stance phase of the painful limb did not change significantly, the single stance phase of right lower limb (the painful side) shortened and the double stance phase prolonged. We think this is because the stance phase of the left lower limb (painless side) increased. A longer double stance phase with a longer contact time in the painless lower limb may be a strategy to reduce loading on the painful side.

Regarding iii) the angle of the head and trunk, anterior tilts in the head, thorax, and pelvis segments were observed in global coordinates, but not in the joint angles of the neck and trunk in local coordinates. Namely, anterior trunk tilt occurred originating from the pelvis; however, there were no changes in the relative positions in the head, thorax, and pelvis. To the best of our knowledge, this was the first study to analyze the effect of plantar pain on trunk posture. Specifically, anterior trunk tilt was observed during the right stance phase. The reason for anterior trunk tilt may be due to a strategy of quick load transfer from the right foot (painful side) to the left foot (painless side). Because a large amount of the body’s mass is contained in the head, arms, and trunk¹⁵⁾, anterior trunk tilt may contribute to the forward shift of the center of mass and thus shorten the time for load transfer of right to left.

Regarding the joints of the lower limb, flexions of the left (painless) hip and knee were limited around the right initial contact. In addition, extensions of the right (painful) hip and knee were limited during the right stance phase. The directions of the changes were almost compatible with the results of the previous studies^{7, 8)}. Limiting the painless limb flexions and limiting the painful limb extension may be done to reduce the load of painful side.

A novel finding of our study is anterior trunk tilt in global coordinates. The trunk plays a major role in generating ground reaction forces¹⁶⁾, and anterior trunk tilt in global coordinates is associated with an increase in lumbar extension moment¹⁷⁾. Therefore, plantar pain may be a burden on the trunk. Our motivation for the present study was that foot plantar pain increases the prevalence of low back pain⁶⁾. Our results may explain one of the mechanisms for this fact.

A limitation of our study is that our artificial pain did not exactly mimic the plantar pain of plantar fasciitis. To mimic the pain more exactly, we could have employed more invasive techniques (e.g., saline solution injection into the plantar fascia), which we did not prefer to use for ethical reasons. To induce pain surely but minimally invasively, we chose our methodology. Another limitation of our study is that we only evaluated the immediate effect of experimentally induced pain. Due to ethical reasons, we do not prefer to conduct a study evaluating the long-term effects of experimentally induced pain.

In conclusion, to clarify the effect of foot plantar pain on the trunk, we conducted an experiment that induced experimental pain in the right foot planta. The results showed anterior trunk tilt during the right stance phase. This may be a burden on the trunk, and one of the possible reasons for increased prevalence of low back pain in the plantar pain patients.

Conflicts of interest

The authors declare no conflicts of interest associated with this manuscript.