Impact of high-heeled and sport shoes on multi-joint external load profile during walking

Abstract

BACKGROUND:

Previous studies have analysed the effect of wearing high-heeled shoes (HHS) on gait analysis, balance and its relation to health. However, further research is needed to study its effect on the difference of chain reactions in the transfer of body impacts from the lower to the upper limbs.

OBJECTIVES:

The aims of the present research were: (a) to compare the effects of wearing HHS on impacts across body joints during walking with sport shoes (SS) as a reference, and (b) to examine such effects at different speeds.

METHODS:

Seven well-trained women completed this study. Incremental treadmill walking test were performed with two different footwear: SS and HHS. Inertial devices were used to quantify the chain reactions at selected anatomical lower limbs and trunk locations. Statistical analysis included the Wilcoxon test with ranges and Cohen’s $d$ effect size with percentage of differences.

RESULTS:

The highest values were found at the heel in both footwear and in both legs (SS: right $=$ 0.76 $\pm$ 0.27, left $=$ 0.79 $\pm$ 0.27; HHS: right $=$ 1.07 $\pm$ 0.38, left $=$ 1.11 $\pm$ 0.41), while the lowest values were registered at lower and upper back. Furthermore, significant differences were found with the HHS load being higher at all locations ( $p<$ 0.05; % ${}_{\textit{diff}}=$ 12.20–36.36%), influenced by the walking speed. In addition, a strong influence of footwear in the change of the laterality profile was found ( $p<$ 0.05).

CONCLUSIONS:

These findings suggest that the use of HHS increase the load on the lower limb and the trunk. Until reaching 5 km/h during walking, no significant differences were found between wearing HHS or SS in accelerometer load, producing exponential differences from this speed. The great between-subject variability implies that within-subject analysis is recommended, as it is more related to real clinical practice.

Keywords

Footwear accelerometry assessment health gait biomechanics

1. Introduction

Gait analysis is well known as the study of human locomotion involving cyclical movements of the body segments [42]. In humans, one of the factors that modifies gait patterns is the use of shoes [2, 23]. Recent evidence has indicated that 78% of women usually wear high-heeled shoes (HHS) [3] and 40% of them do so on a daily basis [4, 46]. Although wearing HHS has been considered to characterise beauty, self-assurance and elegance according to Dai et al. [12], their use has been linked to many medical problems [45].

In this regard, previous research has shown that the use of HHS may cause foot deformities [35] as well as an increase in gastrocnemius strain injuries [11], low back pain [12], ankle sprains [17], knee osteoarthritis [26] or musculoskeletal pain [15]. The identification of the risk factors associated with the occurrence of these injuries is crucial in order to implement preventive strategies. Previous studies have identified that HHS alter the function and position of the ankle joint since the foot is forced to be in plantar flexion [10]. Furthermore, due to the larger breaking forces [40], knee flexion increases [40] and leads to a higher activation of the rectus femoris muscle in order to prevent extra knee flexion rate at heel strike [45]. In addition, plantar flexor muscle activity is decreased by 29% [15], range of motion of the ankle and knee is reduced ( $p<$ 0.05; knee: 3–5 ${}^{\circ}$ ; ankle: 6–9 ${}^{\circ}$ ) [32], stride length and walking speed are reduced [36], transference of weight from the heel to the forefoot is faster [45] and vertical displacements of the centre of mass are higher [2], in comparison with walking in low-heeled shoes. Moreover, it has been reported that during walking, these alterations could increase metabolic cost [14], causing muscle loading and consequently acceleration of muscle fatigue [19]. This could then impair stability and increase the risk of injury [14].

Research on gait analysis has been conducted since the late 19 ${}^{\text{th}}$ century with the availability of video camera systems [18, 25, 29]. The standard gait analysis method has been based on the multi-camera motion capture system and force platforms with the capability of measuring ground-reaction forces [8, 27]. However, this standard gait analysis requires specialised locomotion laboratories, expensive equipment, a lengthy set up and post-data collection processing time. Besides, limitations have been observed in terms of portability, and the freedom to measure in any sport context [1]. To mitigate these problems, an alternative gait analysis method based on wearable sensors, which are much less expensive and can be applied outside the laboratory environment, has been studied and has shown great promise in recent years [41, 42].

However, studies regarding the comparison of the impact between HHS and sport shoes (SS) in several anatomical joints are scarce, and most only consider biomechanical aspects. In addition, to the authors’ knowledge, no previous studies have analysed the chain reactions in the body (forces, accelerations, decelerations, etc.) during walking with accelerometer sensors. The accelerometer load supported by a body structure has been expressed as PlayerLoad (PL ${}_{\text{RT}}$ ) and was validated previously for this purpose [21]. Findings showed that the consequences of using HHS may not affect mainly the foot but also involve the subsequent chain reaction which travels up to the spine [10]. Despite the numerous studies on HHS gait analysis, balance and its relation to health, further research is needed to study how these chain reactions in the body transfer from the lower to the upper limbs, and multiple accelerometers placed in several anatomical locations are suitable tools to perform this analysis. Therefore, the aim of the present study was to compare the effects of wearing HHS on impacts across body joints during walking with SS as a reference. Moreover, a secondary aim was to examine such effects at different speeds. We hypothesize that high-heeled shoes compared to sport ones provoke higher impact at all anatomical locations.

2. Methods

2.1 Design

A cross-sectional study with repeated measures was designed to analyse the multi-joint external workload profile considering different footwear (HHS vs SS). The STROBE guidelines (quality checklist of experimental designs) were followed [7]. All women visited the laboratory for three sessions. The first session was used as a familiarization session and anthropometrical measurements were performed. Then, the second and third sessions were used to collect data from walking on HHS and SS following a randomised design. The testing sessions lasted for about an hour each. In addition, PL ${}_{\text{RT}}$ was assessed through inertial measurement units placed on eight anatomical locations (two units at the heels, two units at the ankles, two units at the knees, one unit at the centre of mass and another one at the scapulae) in order to measure external workload.

2.2 Participants

Seven well-trained women participated voluntarily in this research (age: 19.1 $\pm$ 0.45 years; height: 167.81 $\pm$ 4.22 cm; body mass: 76.16 $\pm$ 5.78 kg). All the participants met the following criteria: (i) no history of orthopaedic disorders in the lower legs or injuries that could have affected testing; (ii) no abnormal gait pattern or discomfort during ambulation. Before the start of this investigation, all subjects were fully informed about the testing protocol and their written informed consent was obtained. The procedure was approved by the Institutional Ethics Review Committee of the University (Registration Code 2061–2018), which is in consistent with the code of ethics of the World Medical Association (Declaration of Helsinki, 7th edition).

2.3 Equipment and measurement

2.3.1 Anthropometric characteristics

Each participant’s height and body mass were assessed. Specifically, height was measured to the nearest 0.5 cm during a maximal inhalation using a wall-mounted stadiometer (SECA, Hamburg, Germany). Body mass was obtained with an 8-electrode segmental body composition monitor BC-601 model (TANITA, Tokyo, Japan).

2.3.2 External load

A WIMU PRO ${}^{\text{TM}}$ inertial device (RealTrack Systems, Almeria, Spain) was used to quantify the chain reactions (forces, accelerations, decelerations, etc.) at selected anatomical sites during walking. This device contains four triaxial accelerometers that can detect and measure movement using a micro-electromechanical system with an adjustable sampling frequency from 10 to 1000 Hz. The full-scale output ranges are $\pm$ 16 g, $\pm$ 16 g, $\pm$ 32 g and $\pm$ 400 g. Each device has its own microprocessor, 2-GB flash drives and high-speed USB interface, to record, store and upload data. The device is powered by an internal battery with 4-h of life, weighs 70-g and is 81 $\times$ 45 $\times$ 16 mm in size. In this case, accelerometers were set at 1000 Hz.

The external workload variable registered was PlayerLoad by RealTrack Systems (PL ${}_{\text{RT}}$ ). This is an accelerometer-derived measurement of total body load in its 3 axes (vertical, anterior-posterior and medial-lateral) [21, 37]. It is represented in arbitrary units (a.u.) and is calculated from the following equation at a 1000 Hz sampling frequency where: $PL_{n}$ is the player load calculated in the current instant; $n$ is the current instant in time; $n-1$ is the previous instant in time; $X_{n},Y_{n}$ and $Z_{n}$ are the values of Body Load for each axis of movement in the current instant in time; $X_{n-1},Y_{n-1}$ and $Z_{n-1}$ are the values of Body Load for each axis of movement in the previous instant in time.

$\displaystyle PL_{n}=\sqrt{\frac{\begin{array}[]{c}({X_{n}-X_{n-1}})^{2}+({Y_{% n}-Y_{n-1}})^{2}\\ +({Z_{n}-Z_{n-1}})^{2}\\ \end{array}}{1000}}$ $\displaystyle\textit{Accumulated }PL=\sum_{n=0}^{m}PL_{n}\times 0.01$

For this purpose, eight inertial devices were placed on each subject at the following anatomical locations: (a) right heel, (b) left heel, (c) right ankle, (d) left ankle, (e) right knee, (f) left knee, (g) lower back and (h) upper back. First, the heel devices were placed on the heel-counter in SS and on the heel in HHS. At the knee and ankle, the inertial devices were located on the outside of the leg in all participants, placing them at the ankle 3 centimetres from the beginning of the malleolus [28] and at the knee 3 centimetres from the horizontal direction of the kneecap [41]. The device placed on the lower back was over the L4 vertebra [31], while the device placed on the upper back was over the interscapular line (vertebrae T5-T7) [22].

Figure 1.

Flowchart of the study procedures.

Figure 2.

Device placement on one of the participants during the investigation. A. On the left: participant performing the test with HHS; B. On the right: participant performing the test with SS.

These devices were attached to the participants with a specifically designed adhesive elastic band, except the upper back device that was placed in a specially designed harness (Fig. 1) and the heel devices that were placed using stickers. Prior to placement, the inertial devices were calibrated manually, according to the manufacturer’s recommendations, and synchronised. This process eliminated four 3D accelerometer error sources: offset error, scaling error, non-orthogonal error and random error [43]; and the static and dynamic bias of the raw data between inertial devices have been assessed previously, obtaining excellent results [20].

2.3.3 Footwear

During the study, participants wore two types of footwear: (a) high-heel shoes (HHS) with 80-mm heel and (b) sport shoes (SS) with 10-mm drop.

2.4 Procedure

Firstly, a non-probabilistic convenience sampling was performed to select participants among female students of the Physical Activity and Sport Science degree of the University of Murcia (Spain). Data collection was realized in the laboratory of Sport Science Faculty in three different sessions between 2 and 6 April 2018. The same two researchers collected all data during the assessments. The first session consisted of a familiarisation with the testing procedure and experimental equipment, and measurement of the anthropometric characteristics. During the second and third session, an incremental treadmill walking test with different type of footwear (HHS or SS) was performed randomly, being both sessions at the same time of the day for each participant (between 8 and 11 a.m.) to avoid the effect of circadian rhythm (Fig. 2).

Before starting the experimental protocol, participants carried out a standardised warm-up consisting of 3 minutes of low intensity walking at 3 km/h. The warm-up period was monitored in real time with SPRO ${}^{\text{TM}}$ software to verify the perfect functioning of the devices. Then, the incremental treadmill walking test started at 3 km/h, and subsequently the velocity was increased by 1 km/h each minute until achieving 6 km/h. The total duration of each test was 4 minutes. The trials were performed by participants wearing HHS and SS in randomised order. Active rest of 10 minutes was taken between the test performances with the HHS and the SS.

Finally, SPRO ${}^{\text{TM}}$ software (RealTrack Systems, Almeria, Spain) was used to extract the inertial device data from each trial. To synchronise the eight inertial devices, they were placed in a waterproof housing and then rotated through 360 ${}^{\circ}$ ; S PRO ${}^{\text{TM}}$ software automatically synchronised the devices for analysis.

2.5 Statistical analysis

Before starting the analysis, the normality test for criteria assumption was performed to determine the hypothesis test model. For this, the Shapiro-Wilk test was performed, showing a non-normal distribution so that non-parametric statistics were used for the analysis [16]. Descriptive statistics are presented as medians and ranges. The Wilcoxon test was used to identify the differences between types of shoes (SS vs. HHS) in the different anatomical locations and in relation to laterality [16]. To determine the magnitude of differences, Cohen’s formula for effect size ( $d$ ) [9] was used and the interpretation of the results was based on the following criteria: insignificant (0–0.2), low (0.2–0.6), moderate (0.6–1.2), high (1.2–2.0) and very high ( $>$ 2.0) [24] with percentage of differences. Data were analysed using the Statistical Package for the Social Sciences (SPSS, release 24.0; SPSS Inc., Chicago IL, USA). Significance was accepted as $p<$ 0.05 level.

3. Results

Descriptive analyses with the dynamics of the PL ${}_{\text{RT}}$ variable for all functional anatomical locations, separated per leg and per types of shoes (SS and HHS), are presented in Table 2 and Fig. 3. The highest values were found at the heel in the two analysed types of shoes and in both legs (SS: right $=$ 0.76, left $=$ 0.79; HHS: right $=$ 1.07, left $=$ 1.08). In contrast, the lowest values were found at the upper back (SS $=$ 0.08; HHS $=$ 0.12) and lower back (SS $=$ 0.11; HHS $=$ 0.18).

Table 2 shows the within and between-subject comparative analysis among type of shoes at all anatomical locations in both legs. Significant statistical differences were found in almost all participants ( $p<$ 0.01), with

Table 1
Descriptive analysis of the PlayerLoad (PL ${}_{\text{RT}}$ ) variable at all anatomical locations and type of shoes

Type of shoes	Participant	Right leg						Left leg						Whole body
		Heel		Ankle		Knee		Heel		Ankle		Knee		Lower back		Upper back
		Median	Range	Median	Range	Median	Range	Median	Range	Median	Range	Median	Range	Median	Range	Median	Range
Sport shoes	1	0.83	0.80	0.55	0.50	0.41	0.46	0.83	0.80	0.47	0.40	0.39	0.38	0.15	0.14	0.09	0.12
	2	0.81	0.78	0.43	0.37	0.39	0.45	0.81	0.78	0.44	0.35	0.36	0.38	0.10	0.11	0.08	0.09
	3	0.87	0.89	0.57	0.52	0.40	0.37	0.87	0.89	0.51	0.46	0.39	0.38	0.15	0.15	0.10	0.009
	4	0.68	0.75	0.42	0.40	0.35	0.42	0.68	0.75	0.37	0.37	0.33	0.39	0.08	0.08	0.15	0.16
	5	0.51	0.69	0.31	0.45	0.26	0.41	0.51	0.69	0.30	0.36	0.25	0.34	0.10	0.14	0.05	0.08
	6	0.53	0.65	0.33	0.40	0.29	0.40	0.53	0.65	0.29	0.33	0.27	0.33	0.10	0.12	0.05	0.08
	7	1.14	1.57	0.59	0.59	0.41	0.53	1.14	1.38	0.58	0.55	0.40	0.44	0.14	0.16	0.10	0.17
	Total	0.76	1.27	0.46	0.94	0.38	0.63	0.79	1.67	0.44	0.83	0.36	0.55	0.11	0.19	0.08	0.20
High-heeled	1	0.99	1.31	0.66	0.59	0.44	0.47	1.03	1.41	0.67	0.62	0.42	0.41	0.17	0.31	0.11	0.33
shoes	2	1.52	1.76	0.73	0.77	0.56	0.65	1.54	1.73	0.64	0.76	0.42	0.46	0.07	0.33	0.15	0.16
	3	1.01	0.76	0.61	0.48	0.38	0.35	1.04	0.91	0.67	0.57	0.44	0.36	0.21	0.18	0.14	0.12
	4	0.69	0.86	0.52	0.77	0.35	0.45	0.71	0.88	0.67	0.94	0.40	0.68	0.11	0.11	0.16	0.27
	5	0.70	0.81	0.49	0.63	0.26	0.33	0.66	0.97	0.44	0.53	0.25	0.34	0.13	0.20	0.09	0.10
	6	0.98	0.97	0.61	0.64	0.41	0.48	0.91	0.90	0.58	0.54	0.35	0.36	0.16	0.20	0.10	0.14
	7	1.57	1.23	0.84	1.05	0.48	0.61	1.68	1.42	0.88	1.06	0.44	0.61	0.25	0.41	0.16	0.22
	Total	1.07	1.94	0.69	1.29	0.41	0.86	1.08	2.15	0.66	1.24	0.39	0.66	0.18	0.51	0.12	0.36

Table 2

Within and between-subject comparison in the PlayerLoad (PL ${}_{\text{RT}}$ ) variable between types of shoes as a function of the location of the tri-axial accelerometer in each leg

Leg	Participant	Heel				Ankle				Knee				Lower back				Upper back
		$Z$	Ranges (ss-hhs-d)	$p$	$d$	$Z$	Ranges (ss-hhs-d)	$p$	$d$	$Z$	Ranges (ss-hhs-d)	$p$	$d$	$Z$	Ranges (ss-hhs-d)	$p$	$d$	$Z$	Ranges (ss-hhs-d)	$p$	$d$
Right	1	$-$ 2.40	7-12-1	0.01	$-$ 0.45	$-$ 2.69	5-15-0	0.01	$-$ 0.78	$-$ 3.92	0-20-0	0.00	$-$ 0.45	$-$ 3.94	0-20-0	0.00	$-$ 0.86	$-$ 2.46	3-15-2	0.01	$-$ 0.61
	2	$-$ 3.92	0-20-0	0.00	$-$ 1.68	$-$ 3.92	0-20-0	0.00	$-$ 1.45	$-$ 3.92	0-20-0	0.00	$-$ 0.70	$-$ 1.82	11-9-0	0.07	0.56	$-$ 3.94	0-20-0	0.00	$-$ 1.53
	3	$-$ 3.92	0-20-0	0.00	$-$ 0.85	$-$ 3.85	1-19-0	0.00	$-$ 0.61	$-$ 2.34	8-11-1	0.02	$-$ 0.15	$-$ 3.92	0-20-0	0.00	$-$ 1.26	$-$ 3.92	0-20-0	0.00	$-$ 1.12
	4	$-$ 2.87	3-16-1	0.00	$-$ 0.27	$-$ 3.92	0-20-0	0.00	$-$ 0.81	$-$ 2.19	3-15-2	0.03	$-$ 0.13	$-$ 3.96	0-20-0	0.00	$-$ 0.84	$-$ 3.37	1-17-2	0.00	$-$ 0.43
	5	$-$ 3.89	0-19-1	0.00	$-$ 0.78	$-$ 3.92	0-20-0	0.00	$-$ 1.06	$-$ 0.63	8-7-5	0.53	0.09	$-$ 3.83	0-19-1	0.00	$-$ 0.77	$-$ 3.87	0-19-1	0.00	$-$ 0.86
	6	$-$ 3.92	0-20-0	0.00	$-$ 1.18	$-$ 3.92	0-20-0	0.00	$-$ 1.23	$-$ 3.71	2-17-1	0.00	$-$ 0.34	$-$ 3.42	1-15-4	0.00	$-$ 0.77	$-$ 3.75	0-18-2	0.00	$-$ 1.00
	7	$-$ 3.92	0-20-0	0.00	$-$ 0.89	$-$ 3.92	0-20-0	0.00	$-$ 0.95	$-$ 3.83	0-19-1	0.00	$-$ 0.35	$-$ 3.92	0-20-0	0.00	$-$ 1.02	$-$ 3.92	0-20-0	0.00	$-$ 0.70
	Total	$-$ 9.87	10-128-2	0.00	$-$ 0.92	$-$ 10.07	6-134-0	0.00	$-$ 0.98	$-$ 8.16	21-109-10	0.00	$-$ 0.33	$-$ 7.28	12-123-5	0.00	$-$ 0.71	$-$ 9.72	4-129-7	0.00	$-$ 0.84
Left	1	$-$ 0.30	6-13-1	0.76	$-$ 0.73	$-$ 3.73	0-18-2	0.00	$-$ 0.49	$-$ 2.51	6-12-2	0.01	$-$ 0.17
	2	$-$ 3.92	0-20-0	0.00	$-$ 1.66	$-$ 3.92	0-20-0	0.00	$-$ 1.20	$-$ 3.92	0-20-0	0.00	$-$ 0.38
	3	$-$ 3.92	0-20-0	0.00	$-$ 0.86	$-$ 3.92	0-20-0	0.00	$-$ 1.28	$-$ 3.92	0-20-0	0.00	$-$ 0.67
	4	$-$ 1.06	8-8-4	0.29	$-$ 0.08	$-$ 3.92	0-20-0	0.00	$-$ 1.39	$-$ 3.92	0-20-0	0.00	$-$ 0.58
	5	$-$ 3.82	0-19-1	0.00	$-$ 0.76	$-$ 3.92	0-20-0	0.00	$-$ 1.07	$-$ 1.62	3-10-7	0.11	$-$ 0.10
	6	$-$ 3.92	0-20-0	0.00	$-$ 1.00	$-$ 3.92	0-20-0	0.00	$-$ 1.38	$-$ 3.14	2-17-1	0.00	$-$ 0.26
	7	$-$ 3.92	0-20-0	0.00	$-$ 0.89	$-$ 3.92	0-20-0	0.00	$-$ 1.17	$-$ 3.84	1-19-0	0.00	$-$ 0.37
	Total	$-$ 9.35	14-120-6	0.00	$-$ 0.91	$-$ 10.19	0-138-2	0.00	$-$ 1.16	$-$ 8.79	15-114-11	0.00	$-$ 0.37

Figure 3.

PlayerLoad (PL ${}_{\text{RT}}$ ) dynamics wearing sport shoes and high-heeled shoes at all anatomical locations. ${}^{*}$ Statistical differences ( $p<$ 0.05).

Table 3

Within and between-subject comparison in the PlayerLoad (PL ${}_{\text{RT}}$ ) variable between legs as a function of the location of the tri-axial accelerometer and type of shoes

Participant	Heel				Ankle				Knee
	$Z$	Ranges (l-r-d)	$p$	$d$	$Z$	Ranges (l-r-d)	$p$	$d$	$Z$	Ranges (l-r-d)	$p$	$d$
1	$-$ 3.94	20-0-0	0.00	0.04	$-$ 0.61	9-10-1	0.55	0.19	$-$ 3.04	4-14-2	0.00	$-$ 0.23
2	$-$ 3.41	15-4-1	0.00	$-$ 0.22	$-$ 0.70	8-9-3	0.49	0.00	$-$ 3.74	0-18-2	0.00	0.30
3	$-$ 1.46	11-7-2	0.14	$-$ 0.11	$-$ 3.83	0-19-1	0.00	0.30	$-$ 0.58	6-8-6	0.56	0.08
4	1.74	12-6-2	0.08	$-$ 0.13	$-$ 3.95	0-20-0	0.00	0.30	$-$ 2.71	2-13-5	0.01	0.14
5	1.80	6-11-3	0.07	0.10	$-$ 3.84	0-19-1	0.00	0.23	$-$ 2.33	4-11-5	0.02	0.08
6	$-$ 0.92	13-7-0	0.36	0.00	$-$ 3.64	0-17-3	0.00	0.33	$-$ 3.74	0-18-2	0.00	0.24
7	$-$ 2.33	15-5-0	0.02	$-$ 0.23	$-$ 3.19	2-15-3	0.00	0.15	$-$ 2.68	5-12-3	0.01	0.17
Total	$-$ 4.41	92-40-8	0.00	$-$ 0.10	$-$ 7.89	19-109-12	0.00	0.22	$-$ 7.46	21-94-25	0.00	0.12
1	$-$ 2.14	2-12-6	0.03	$-$ 0.20	$-$ 2.58	15-5-0	0.01	0.50	$-$ 3.92	15-4-1	0.00	0.10
2	$-$ 0.90	13-7-0	0.37	$-$ 0.04	$-$ 3.27	2-18-0	0.00	0.26	$-$ 3.92	0-20-0	0.00	0.69
3	$-$ 2.88	17-2-1	0.00	$-$ 0.18	$-$ 3.83	19-0-1	0.00	$-$ 0.33	$-$ 3.88	19-1-0	0.00	$-$ 0.42
4	$-$ 0.64	9-11-0	0.52	0.07	$-$ 3.92	20-0-0	0.00	$-$ 0.44	$-$ 3.92	20-0-0	0.00	$-$ 0.36
5	$-$ 1.92	4-14-2	0.06	0.07	$-$ 3.77	1-18-1	0.00	0.28	$-$ 1.35	7-6-7	0.18	$-$ 0.10
6	$-$ 3.68	2-18-0	0.00	0.20	$-$ 3.69	1-18-1	0.00	0.20	$-$ 3.53	1-16-3	0.00	0.37
7	$-$ 3.75	19-1-0	0.00	$-$ 0.30	$-$ 2.95	16-3-1	0.00	$-$ 0.10	$-$ 3.45	1-15-4	0.00	0.14
Total	$-$ 1.69	66-65-9	0.12	$-$ 0.09	$-$ 2.25	74-62-4	0.02	$-$ 0.02	$-$ 3.46	48-78-14	0.00	0.08

Note. M: mean; SD: standard deviation; $Z$ : Z-value of Wilcoxon test; Ranges: values distribution (r: right leg, l: left leg, d: draws); $p$ : $p$ -value; $d$ : Cohen’s d effect size.

the HHS load being higher ( $d=$ 1.16–0.33) than SS at all locations.

Finally, the within and between-subject comparative analysis between legs at all anatomical locations in both types of shoes is presented in Table 3. The data show significantly different external load profiles in relation to laterality ( $p<$ 0.05), with a strong influence of type of shoes in the change of the laterality profile. In the specific analysis in each location, while with SS the PL ${}_{\text{RT}}$ was significantly higher in the left heel ( $p<$ 0.05; $d=-$ 0.10), conversely, when HHS were worn there were not significant differences in laterality ( $p=$ 0.12; $d=-$ 0.09). In the ankle, differences were found for the right ankle with SS ( $p<$ 0.01; $d=$ 0.22) and for the left ankle in HHS ( $p=$ 0.02; $d=-$ 0.02). Finally, at the knee, no differences in laterality profile were found, PL ${}_{\text{RT}}$ being higher in the right knee (SS: $p<$ 0.01; $d=$ 0.12; HS: $p<$ 0.01; $d=$ 0.08).

4. Discussion and implications

Previous studies have shown that the effects of HHS are not restricted to the feet as they travel from the lower legs to the spine [10]. However, studies regarding the impact of using different types of shoes on the anatomical joints are scarce. A recent review carried out an analysis of the reliability of an accelerometer sensor to monitor neuromuscular load according this multi-joint approach [20]. Thus, the aim of the present study was to compare the effects of wearing HHS or SS on the impacts at specific body locations while walking, and to examine the respective effects at different speeds.

According to the current results, participants who walked with HHS showed significantly higher PL ${}_{\text{RT}}$ ( $p<$ 0.05; % ${}_{\textit{diff}}=$ 12.20–36.36%) compared with SS in all anatomic locations. Furthermore, the present study observed moderate differences in the heel and ankle ( $d=$ 0.91–1.16; % ${}_{\textit{diff}}=$ 28.82–36.36%) and in the lower back and upper back ( $d=$ 0.71–0.84; % ${}_{\textit{diff}}=$ 29.41–35.71%). Conversely, low differences were found in the knee ( $d=$ 0.33–0.37). In addition, our results showed significant differences ( $p<$ 0.01; % ${}_{\textit{diff}}=$ 26.25 $\pm$ 4.51%) between HHS and SS influenced by the walking speed (up-to-4.5 km/h) in the heel, ankle, lower back and upper back.

The location of the accelerometer is very important depending on the measurement aim. Nedergaard et al. [33] performed a correlation analysis between the acceleration of the lower back with the rest of the body segments (scapulae, knee and tibia), showing a low correlation. Their main finding was that body-worn accelerometers only measure the acceleration of the segment they are attached to. Thus, this research concluded that it is complex to measure the acceleration of the whole body by multi-segment movements during sports actions (45 ${}^{\circ}$ and 90 ${}^{\circ}$ changes of direction and straight-line running at 2, 3, 4 and 5 m/s). For this reason, previous studies have reported acceleration in different anatomical locations simultaneously [42]. Özdemir [34] placed the devices at the head, chest, waist, right-wrist, right-thigh and right-ankle to evaluate the fall risk in elderly people. Boersma et al. [6] used five accelerometers on different parts of the waist to measure physical activity; however, a multi-joint analysis has never been performed with the same aim as the present study.

Concerning the effect at different speeds, the results of the present study show that there is a significant relationship between speed and PL ${}_{\text{RT}}$ ( $r=$ 0.88–0.97). This relationship is consistent with those of Barrett et al. [5] and Nedergaard et al. [33] in an incremental running protocol. These investigations studied the accelerometer load in scapulae, COM, pelvis, and tibia, and found a direct relationship between speed and accelerometer load. Besides, this can be seen in Fig. 2 where no significant differences were recorded in any location of the accelerometer before exceeding 4.5 km/h. That is, at the kinematic level, it is possible to reduce the reaction force against the ground and the impacts received by the structures of the body walking at speeds no higher than 4.5–5 km/h. These differences were not found until exceeding 5 km/h in the ankle and 6 km/h in the knee. This could imply a threshold of gait speed for reaction force/impact reduction; although, this study cannot indicate that at the biomechanical level there are no harmful maladjustments for health.

Concerning the effect of wearing a type of footwear, few studies have analysed this phenomenon. Several studies showed higher acceleration when wearing minimalist shoes with respect to sport shoes [38, 39]. Previous studies have investigated this phenomenon from different perspectives. Wang et al. [44] described the biomechanical alterations that were produced by wearing HHS, finding differences in dorsiflexion, plantar flexion and abduction in the forefoot respect to walking barefoot. This alteration produced negative effects in gait biomechanics, not influenced by the heel height (6-to-12 cm) [13]. Thus, this investigation seems to confirm that wearing HHS produces a mismatch of the gait pattern of walking, being more pronounced in the foot and back (moderate differences). According to the results of the present study, there is an increase in the accelerometer load in the structures where there is a modification of the gait biomechanics, being significant at the foot, heel and ankle [44] and at the lower and upper back [12]. For example, a previous study indicated that the pressure exerted on the forefoot increased when wearing HHS, which may lead to significant changes in foot morphology [35]. Specifically, this study found that women wearing HHS had flatter feet and an increased laterally flexed hallux compared to women who wore flat-soled shoes or low-heeled shoes [35]. In addition, HHS was associated with the incidence of varus deformity of the fifth toe [35]. However, at the knee just small differences were found ( $d=$ 0.33–0.37). These differences between ankle and knee are due to a different reduction of range of movement ( $p<$ 0.05; knee: 3–5 ${}^{\circ}$ ; ankle: 6–9 ${}^{\circ}$ ), in addition to the lower range of movement and fewer modification possibilities between the knee and the rest of the joints analysed [32]. Even so, the results obtained in this regard are controversial. Baaklini et al. [3] found no significant differences in lumbar lordosis in people walking in HHS compared to barefoot, concluding that the low back pain that heel users manifested must be associated with other factors. However, other authors have found that HHS influence the movement of the pelvis and spine when walking [12], inducing moderate differences in the external load supported by the lower and upper back.

On the other hand, considering the biomechanical modifications that wearing heels can entail, the results of this study also show a tendency of change in the asymmetry of stepping (Table 3). While wearing SS, the PL ${}_{\text{RT}}$ was significantly higher in the left heel than right heel, and when the same subjects wore HHS, there were no significant differences between left or right heel. This change was more relevant in the ankle, in which differences were found for the right ankle with SS and for the left ankle in HHS. However, this tendency was not observed for the knee. This observation should be viewed with caution due to the small number of subjects, and due to the intra-subject analysis, in some of them this phenomenon was not observed. However, it could indicate a change in the laterality of stepping due to the biomechanical modifications that wearing HHS entails [10].

Previous studies have not successfully found the heel height threshold that would not significantly influence the gait pattern [30, 40]. However, they established the relationship between the heel height and the lumbar flexion angle. In this respect, the authors did not take into account the walking speed, analysing this variable in running and in specific sport skills with sport shoes [5, 33].

While the results of this study have provided information regarding the impact loading with different shoes (HHS vs. SS), some limitations to the study must be acknowledged. One of the limitations in this study concerns the sample studied; it would be interesting to extend this study to include more participants. It would also be interesting for futures studies to analyse the same parameters on a normal surface and not on a treadmill, complementing gait analysis with biomechanical analysis techniques. In addition, since PL ${}_{\text{RT}}$ is an indicator of the load supported by a body structure, this study analysed the impacts absorption across body joints when walking with HHS in order to explain the impact that the type of shoe might have on multi-joint external load profile. However, this may be considered as a limitation of the study since this variable (i.e., PL) may not be representative of other biomechanics-related parameters in human walking pattern.

5. Conclusions

From the results obtained, we can propose three considerations about the use of high-heeled shoes in relation to external load during walking:

•

High-heeled shoes increased the load at all anatomical locations, being the greatest load at the heel and ankle.

•

Until reaching 5 km/h, no significant differences were found between wearing high-heeled shoes and sport shoes in accelerometer load. However, greater speeds are related to exponential differences in impact between shoes.

•

The great between-subject variability implies that within-subject analysis is recommended, as it is more related to real clinical practice.

Footnotes

Conflict of interest

None of the authors have any conflicts of interest to declare.

Funding

The authors Carlos D. Gómez Carmona and José M. Oliva Lozano were supported by a grant from the Spanish Ministry of Innovation, Science and Universities (FPU17/00407 and FPU18/04434). This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Tao

Liu

Zheng

Feng

. Gait analysis using wearable sensors. Sensors. 2012; 12(2): 2255–83. doi: 10.3390/s120202255.

Annoni

Mapelli

Sidequersky

Zago

Sforza

. The effect of high-heeled shoes on overground gait kinematics in young healthy women. Sport Sci Health. 2014; 10: 149–57. doi: 10.1007/s11332-014-0191-z.

Hong

W-H

Lee

Y-H

Lin

Y-H

Tang

SFT

Chen

H-C

. Effect of shoe heel height and total-contact insert on muscle loading and foot stability while walking. Foot Ankle Int. 2013; 34(2): 273–81. doi: 10.1177/1071100712465817.

Baaklini

Angst

Schellenberg

Hitz

Schmid

Tal

, et al. High-heeled walking decreases lumbar lordosis. Gait Posture. 2017; 55: 12–4. doi: 10.1016/j.gaitpost.2017.03.035.

Barkema

Derrick

Martin

. Heel height affects lower extremity frontal plane joint moments during walking. Gait Posture. 2012; 35(3): 483–8. doi: 10.1016/j.gaitpost.2011.11.013.

Yoon

J-Y

D-H

Yoo

W-G

Kwon

Y-R

. Differences in activities of the lower extremity muscles with and without heel contact during stair ascent by young women wearing high-heeled shoes. J Orthop Sci. 2009; 14(4): 418–22. doi: 10.1007/s00776-009-1351-x.

Dai

Zhou

Luo

Zhou

. High-heeled-related alterations in the static sagittal profile of the spino-pelvic structure in young women. Eur Spine J. 2015; 24(6): 1274–81. doi: 10.1007/s00586-015-3857-6.

Wiedemeijer

Otten

. Effects of high heeled shoes on gait. A review. Gait Posture. 2018; 61: 423–30. doi: 10.1016/j.gaitpost.2018.01.036.

Puszczalowska-Lizis

Da̧browiecki

Jandziś

Żak

. Foot deformities in women are associated with wearing high-heeled shoes. Med Sci Monit Int Med J Exp Clin Res. 2019; 25: 7746. doi: 10.12659/MSM.917983.

10.

Cronin

Barrett

Carty

. Long-term use of high-heeled shoes alters the neuromechanics of human walking. J Appl Physiol. 2012; 112(6): 1054–8. doi: 10.1152/japplphysiol.01402.2011.

11.

Foster

Blanchette

Chou

Y-C

Powers

. The influence of heel height on frontal plane ankle biomechanics: implications for lateral ankle sprains. Foot Ankle Int. 2012; 33(1): 64–9. doi: 10.3113/FAI.2012.0064.

12.

Kerrigan

Karvosky

Lelas

Riley

. Men’s shoes and knee joint torques relevant to the development and progression of knee osteoarthritis. J Rheumatol. 2003; 30(3): 529–33.

13.

Esenyel

Walsh

Walden

Gitter

. Kinetics of high-heeled gait. J Am Podiatr Med Assoc. 2003; 93(1): 27–32. doi: 10.7547/87507315-93-1-27.

14.

Cronin

. The effects of high heeled shoes on female gait: a review. J Electromyogr Kinesiol. 2014; 24(2): 258–63. doi: 10.1016/j.jelekin.2014.01.004.

15.

Stefanyshyn

Nigg

Fisher

O’Flynn

Liu

. The influence of high heeled shoes on kinematics, kinetics, and muscle EMG of normal female gait. J Appl Biomech. 2000; 16(3): 309–19. doi: 10.1123/jab.16.3.309.

16.

Mika

Oleksy

Mika

Marchewka

Clark

. The influence of heel height on lower extremity kinematics and leg muscle activity during gait in young and middle-aged women. Gait Posture. 2012; 35(4): 677–80. doi: 10.1016/j.gaitpost.2011.12.001.

17.

Rao

Ripa

Lightbourne

. Predictors of walking speed and stride length in high- and low-heeled footwear. Footwear Sci. 2013; 5(3): 179–84. doi: 10.1080/19424280.2013.808707.

18.

Ebbeling

Hamill

Crussemeyer

. Lower extremity mechanics and energy cost of walking in high-heeled shoes. J Orthop Sports Phys Ther. 1994; 19(4): 190–6. doi: 10.2519/jospt.1994.19.4.190.

19.

Gefen

Megido-Ravid

Itzchak

Arcan

. Analysis of muscular fatigue and foot stability during high-heeled gait. Gait Posture. 2002; 15(1): 56–63. doi: 10.1016/S0966-6362(01)00180-1.

20.

Gavrila

Davis

. 3-D model-based tracking of humans in action: a multi-view approach. In Proceedings CVPR IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1996; pp. 73–80. doi: 10.1109/CVPR.1996.517056.

21.

Karaulova

Hall

Marshall

. Tracking people in three dimensions using a hierarchical model of dynamics. Image Vis Comput. 2002; 20: 691–700. doi: 10.1016/S0262-8856(02)00059-8.

22.

Leardini

Chiari

Croce

Cappozzo

. Human movement analysis using stereophotogrammetry. Gait Posture. 2005; 21(2): 212–25. doi: 10.1016/j.gaitpost.2004.05.002.

23.

Casadio

Morasso

Sanguineti

. Direct measurement of ankle stiffness during quiet standing: implications for control modelling and clinical application. Gait Posture. 2005; 21(4): 410–24. doi: 10.1016/j.gaitpost.2004.05.005.

24.

Kim

Eng

. Magnitude and pattern of 3D kinematic and kinetic gait profiles in persons with stroke: relationship to walking speed. Gait Posture. 2004; 20(2): 140–6. doi: 10.1016/j.gaitpost.2003.07.002.

25.

Akhtaruzzaman

Shafie

Khan

MdR

. Gait analysis: systems, technologies, and importance. J Mech Med Biol. 2016; 16(7): 1630003. doi: 10.1142/S0219519416300039.

26.

Takeda

Tadano

Todoh

Morikawa

Nakayasu

Yoshinari

. Gait analysis using gravitational acceleration measured by wearable sensors. J Biomech. 2009; 42(3): 223–33. doi: 10.1016/j.jbiomech.2008.10.027.

27.

Gomez-Carmona

Bastida-Castillo

González-Custodio

Olcina

Pino-Ortega

. Using an inertial device (WIMU PROTM) to quantify neuromuscular load in running: reliability, convergent validity and the influence of type of surface and device location. J Strength Cond Res. 2019; 34(2): 365–373. doi: 10.1519/JSC.0000000000003106.

28.

Brand

. Standards of reporting: the CONSORT, QUORUM, and STROBE guidelines. Clin Orthop Relat Res. 2009; 467(6): 1393–4. doi: 10.1007/s11999-009-0786-x.

29.

Reche-Soto

Cardona-Nieto

Diaz-Suarez

Gomez-Carmona

Pino-Ortega

. AcelT y PlayerLoad

{}^{\text{TM}}

: Dos variables para la cuantificación de la carga neuromuscular. Rev Int Med Cienc Act Física Deporte 2019; Epub: Ahead of Print.

30.

Klassen

Simpson

Lim

Louie

Parappilly

Sakakibara

, et al. “Stepping up” activity poststroke: ankle-positioned accelerometer can accurately record steps during slow walking. Phys Ther. 2016; 96(3): 355–60. doi: 10.2522/ptj.20140611.

31.

Leirós-Rodríguez

Romo-Pérez

García-Soidán

. Validity and reliability of a tool for accelerometric assessment of static balance in women. Eur J Physiother. 2017; 19(4): 243–8. doi: 10.1080/21679169.2017.1347707.

32.

Goodworth

Perrone

Pillsbury

Yargeau

. Effects of visual focus and gait speed on walking balance in the frontal plane. Hum Mov Sci. 2015; 42: 15–26. doi: 10.1016/j.humov.2015.04.004.

33.

Wang

Liu

Fan

. Design and Calibration for a Smart Inertial Measurement Unit for Autonomous Helicopters Using MEMS Sensors, IEEE. 2006; 956–61. doi: 10.1109/ICMA.2006.257754.

34.

Gómez-Carmona

Bastida-Castillo

García-Rubio

Ibáñez

Pino-Ortega

. Static and dynamic reliability of WIMU PROTM accelerometers according to anatomical placement. Proc Inst Mech Eng Part P J Sports Eng Technol. 2018; 233(2): 238–48. doi: 10.1177/2F1754337118816922.

35.

Field

. Discovering Statistics Using IBM SPSS Statistics. 4th ed. London: SAGE. 2013.

36.

Cohen

. A coefficient of agreement for nominal scales. Educ Psychol Meas. 1960; 20: 37–46. doi: 10.1177/001316446002000104.

37.

Hopkins

Marshall

Batterham

Hanin

. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc. 2009; 41(1): 3–13. doi: 10.1249/MSS.0b013e31818cb278.

38.

Nedergaard

Robinson

Eusterwiemann

Drust

Lisboa

Vanrenterghem

. The relationship between whole-body external loading and body-worn accelerometry during team-sport movements. Int J Sports Physiol Perform. 2017; 12(1): 18–26. doi: 10.1123/ijspp.2015-0712.

39.

Özdemir

. An analysis on sensor locations of the human body for wearable fall detection devices: principles and practice. Sensors. 2016; 16(8): 1161. doi: 10.3390/s16081161.

40.

Boerema

van Velsen

Schaake

Tönis

Hermens

. Optimal sensor placement for measuring physical activity with a 3D accelerometer. Sensors. 2014; 14(2): 3188–206. doi: 10.3390/s140203188.

41.

Barrett

Midgley

Lovell

. PlayerLoadTM: reliability, convergent validity, and influence of unit position during treadmill running. Int J Sports Physiol Perform. 2014; 9(6): 945–52. doi: 10.1123/ijspp.2013-0418.

42.

Sinclair

Fau-Goodwin

Richards

Shore

. The influence of minimalist and maximalist footwear on the kinetics and kinematics of running. Footwear Sci. 2016; 8(1): 33–9. doi: 10.1080/19424280.2016.1142003.

43.

Sinclair

Sant

. The effects of cross-fit footwear on the kinetics and kinematics of running. Footwear Sci. 2017; 9(1): 41–8. doi: 10.1080/19424280.2016.1268212.

44.

Wang

Baker

. Analysis of foot kinematics wearing high heels using the oxford foot model. Technol Health Care. 2018; 26(5): 815–23. doi: 10.3233/THC-181264.

45.

Di Sipio

Piccinini

Pecchioli

Germanotta

Iacovelli

Simbolotti

, et al. Walking variations in healthy women wearing high-heeled shoes: shoe size and heel height effects. Gait Posture. 2018; 63: 195–201. doi: 10.1016/j.gaitpost.2018.04.048.

46.

Lee

C-M

Jeong

E-H

Freivalds

. Biomechanical effects of wearing high-heeled shoes. Int J Ind Ergon. 2001; 28(6): 321–6. doi: 10.1016/S0169-8141(01)00038-5.

Impact of high-heeled and sport shoes on multi-joint external load profile during walking

Abstract

BACKGROUND:

OBJECTIVES:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Methods

2.1 Design

2.2 Participants

2.3 Equipment and measurement

2.3.1 Anthropometric characteristics

2.3.2 External load

2.4 Procedure

2.5 Statistical analysis

3. Results

Table 1 Descriptive analysis of the PlayerLoad (PL RT ) variable at all anatomical locations and type of shoes

5. Conclusions

Footnotes

Conflict of interest

Funding

References

Table 1
Descriptive analysis of the PlayerLoad (PL ${}_{\text{RT}}$ ) variable at all anatomical locations and type of shoes