Effects of Orthotic Insoles on Gait Biomechanics in Runners With Flatfoot Under Different Gait Loads

ABSTRACT

The purpose of this study was to quantify the biomechanical performance of the medial longitudinal arch (MLA) in the normal foot (NF), flatfoot (FF), and flatfoot with orthotic insoles (FF + insole) under different gait loads to identify their movement response strategies. Male runners with NF and flexible FF were instructed to walk and run using a rearfoot strike pattern. Subsequently, the FF group underwent the same test again while wearing orthotic insoles. The Qualisys capture system and Kistler force platform were synchronized to collect kinetic and kinematic data. A mixed‐design two‐way ANOVA and one‐dimensional statistical parametric mapping (1‐d SPM) were used to analyze the kinematics and kinetics of MLA and the kinematics of the first metatarsophalangeal joint (1st MTPJ) during the stance phase under different gait loads in the NF, FF, and FF + insole groups. Results showed that gait load and foot type had an interactive effect on MLA compression (p = 0.003). FF exhibited a smaller peak plantarflexion moment of the MLA during walking compared with NF (p = 0.001). The FF group exhibited a smaller dorsiflexion angle of the 1st MTPJ, with dorsiflexion initiating earlier. After wearing orthotic insoles, the kinematics of the MLA and the 1st MTPJ in the FF group shifted closer to those of the NF group, but the kinetics did not change. In conclusion, orthotic insoles modify the kinematics of the foot arch in flatfoot through physical support, but the kinetics remain unchanged compared with those observed without insoles. This indicates that after wearing orthotic insoles, flatfoot runners adjust their foot movement response strategy to maintain their habitual biomechanical pattern.

1. Introduction

The foot, as the starting point of the lower limb kinetic chain, plays a crucial role in daily life. The foot's anatomy is intricate and detailed, with the medial longitudinal arch (MLA) serving as a distinctive arch‐like structure. During gait, the MLA and the plantar fascia work together to transmit forces and cycle energy, effectively absorbing impact and propelling human movement [1, 2]. During human growth and development, the collapse and disappearance of the MLA can lead to the development of flatfoot, with an incidence rate ranging from 13.60% to 26.62% [3, 4]. In the past, flatfoot was considered an abnormal foot type that compromised foot health and athletic performance. However, recent studies suggest that, apart from medial tibial stress syndrome, patellofemoral pain syndrome, and non‐specific lower limb overuse injuries, flatfoot is not associated with other diseases. Relying solely on foot type to predict injury risk is unreliable [5]. Academic understanding of flatfoot is shifting from viewing it as a “foot disorder” to recognizing it as a “healthy anatomical characteristic” [6]. However, excessive pronation in flatfoot, accompanied by MLA collapse or disappearance, rearfoot eversion, and forefoot abduction, can still lead to biomechanical differences between the lower limbs and those of a normal foot [7, 8, 9]. Therefore, the value of studying flatfoot today lies not in determining whether it is pathological, but in understanding how it influences lower limb biomechanics in various athletic contexts and whether it may pose a potential risk factor.

In different gait tasks, such as transitioning from walking to running, the increase in stride velocity and ground impact forces leads to a higher load on the arch during gait [10]. Using a multi‐segment foot model to explore the kinematic behavior of the normal foot during walking and running, it was found that the MLA angle significantly changed with variations in load, and the kinematic coupling between the foot segments also differed [11]. Kinetic analysis revealed greater midtarsal joint positive work during rearfoot strike under the running condition compared with walking [12]. This suggests that the normal arch exhibits distinct movement response strategies under different gait loading conditions. However, various characteristics, including tibial internal rotation, heel eversion, and MLA flattening, have altered the movement patterns of flatfoot runners' feet [13, 14]. It remains unclear whether variations in gait loading affect the movement response strategies of flatfoot, an understanding of which is crucial for determining whether relevant orthotic devices can effectively function across various movement states.

Arch‐support orthotic insoles are a common intervention used to modify the lower limb biomechanics during flatfoot gait. They can improve plantar pressure distribution, restrict excessive foot motion, and optimize lower limb kinematic patterns, thereby enhancing comfort and reducing the risk of sports‐related injuries [6, 15, 16, 17]. However, some studies have suggested that wearing arch‐support orthotic insoles may limit foot joint mobility [18] and have negative effects on lower limb joints [19]. Notably, a marked increase in knee varus moment has been observed during high‐load tasks such as running. Furthermore, previous studies have shown that arch‐supporting insoles can significantly alter plantar pressure distribution in the midfoot during running, typically resulting in reduced pressure in this region [20]. Although such findings reflect changes in foot‐ground contact patterns, they fail to elucidate the midfoot joint's biomechanical response to increased gait loads. The absence of research on the three‐dimensional dynamics of the midfoot—such as MLA moments—has limited a deeper understanding of the mechanisms through which orthotic insoles exert their effects.

Despite widespread academic attention to flatfoot, the motion response strategies of the MLA—considered a core functional unit of the foot—under varying gait loads remain poorly understood, and its kinetic characteristics warrant further investigation. Therefore, in this study, we quantified the biomechanical performance of the lower limbs and foot under different gait loads in the normal foot, flatfoot, and flatfoot with orthotic insoles, with the aim of identifying their movement response strategies and providing theoretical insights into the dynamic regulatory effects of arch‐support orthotic insoles. We hypothesized that, as gait load increased, the MLA angle compression in flatfoot runners would be smaller than in normal‐foot runners, and that wearing orthotic insoles would enable flatfoot runners to exhibit an MLA movement pattern more closely resembling that of normal‐foot runners.

2. Materials and Methods

2.1. Participants

Based on effect sizes reported in the previous similar study [21], we extracted the partial η ² values for both main and interaction effects of influencing factors on foot kinematic and kinetic parameters, using their median values for estimation. The lowest partial η ² was approximately 0.073, corresponding to a Cohen's d of around 0.28. With a statistical power set at 80% and a significance level of 0.05, the minimum required sample size was calculated to be 20 participants using G*Power. A total of 25 male runners were ultimately recruited: 14 with normal foot, all of whom successfully completed the experiment (NF group, n = 14, age: 21.50 ± 2.20 years, height: 1.75 ± 0.03 m, body mass: 68.00 ± 4.17 kg, AHI: 0.34 ± 0.01); and 11 with flatfoot, among whom 10 completed the experiment successfully, while one participant withdrew due to personal reasons (FF group, n = 10, age: 20.50 ± 1.02 years, height: 1.79 ± 0.05 m, body mass: 71.10 ± 5.56 kg, AHI: 0.29 ± 0.01). All participants were right‐leg dominant rearfoot strike runners. Before the experiment, the participants' foot type was screened based on the arch height index (AHI) and hallux valgus angle (HVA). Feet with the AHI ≤ 0.31 were identified as flatfoot, while those with 0.31 < AHI ≤ 0.37 and the HVA ≤ 20° were defined as normal feet [22, 23]. The inclusion of HVA ensured that participants with hallux valgus deformity were excluded from the normal foot group. The following criteria were used for participant selection: (i) weekly running distance ≥ 20 km; (ii) no lower limb injuries or neurological disorders in the past 6 months; and (iii) no intense physical activity within 24 h before testing. This study was approved by the Institutional Review Board of Shanghai University of Sport (No. 102772024RT011). Informed consent was obtained from all participants before the formal experiment.

2.2. Data Collection

We used the 10‐camera Qualisys motion capture system (Qualisys, Sweden, sampling frequency: 200 Hz) in synchrony with the Kistler three‐dimensional force platform (Kistler, Switzerland, sampling frequency: 1000 Hz) to collect kinematic and kinetic data. Additionally, we employed the Smartspeed split timing and real‐time feedback system (VALD, Australia) to control the walking and running speeds of the participants.

2.3. Procedures

After the participants completed their warm‐up, the experimenter placed markers on the participants' feet according to the IOR foot model. These markers were applied to the following anatomical landmarks: left/right lateral malleolus, left/right medial malleolus, left/right posterior calcaneal prominence, left/right medial calcaneal prominence, left/right lateral calcaneal prominence, left/right first metatarsal head, right navicular tuberosity, right first metatarsal base, left/right hallux, right second metatarsal head, right second metatarsal base, left/right fifth metatarsal head, and right fifth metatarsal base (Figure 1) [24]. The NF and the FF groups performed walking tests at a speed of 1.2 m/s ± 10% [22], and running tests at a speed of 3.33 m/s ± 10% [25], while wearing experimental shoes (Lining, AGLT137, China, Figure 1; Table 1). Subsequently, the FF group was required to wear full‐length orthotic insoles with arch support (Stabli, China; Figure 1; Table 1), and walking and running tests were conducted again at the aforementioned speeds. Walking and running speeds were monitored within a designated testing zone, which was defined as the area 1 m before and 1 m after the force plate, and the average velocity within this range was used as the participant's trial speed. The distance from the starting point to the testing zone was approximately 3 m. A 3‐min rest was provided between each test.

FIGURE 1.

(A) to (D) indicate the marker placement positions on the right foot, while (E) represents the arch‐support orthotic insole used in this study.

TABLE 1.

Heights and hardness levels in the forefoot, arch support, and hindfoot regions of the orthotic insoles and experimental shoes.

Variables	Height (mm)			Hardness (shore C)
	Forefoot	Arch support	Hindfoot	Forefoot	Arch support	Hindfoot
Experimental shoe	20.0	30.0	30.0	73.0	80.0	66.5
Orthotic insole	4.5	25.0	16.5	59.5	50.0	59.5

2.4. Data Reduction

A complete stance phase following the contact of the dominant foot with the force plate was extracted and preprocessed—including marker labelling and gap filling—using Qualisys Track Manager software. The processed data were then imported into the Visual 3D software to construct a multi‐segment foot model. Kinematic parameters, including the angle of the MLA and the relative motion between the hallux and forefoot, were computed. All data were low‐pass filtered using a 7 Hz cutoff frequency. The moment when the force on the force platform reached 20 N was marked as the heel strike, and the moment when the force began to drop below 20 N was marked as the toe‐off [26]. The interval between heel strike and toe‐off was defined as the stance phase and was subsequently time‐normalized. Additionally, kinetic data were normalized to body weight.

2.5. Data Analysis

2.5.1. MLA Angle (${\theta }_{\mathrm{MLA}}$)

In this study, ${\theta }_{\mathrm{MLA}}$ represents the three‐dimensional angle between two virtual vectors in a three‐dimensional coordinate system, where the two vectors are $\overrightarrow{\mathit{NS}}$ and $\overrightarrow{\mathit{NM}}$, as shown in Figure 2. $\overrightarrow{\mathit{NS}}$ is a vector of the virtual segment from the navicular tuberosity (FNT) to the medial aspect of the calcaneus (FST), and $\overrightarrow{\mathit{NM}}$ is a vector of the virtual segment from the navicular tuberosity to the first metatarsal head (FM1) [7].

FIGURE 2.

The landmarks required to calculate the MLA angle.

2.5.2. MLA Angle Compression ($\mathrm{\Delta }{\theta }_{\mathrm{MLA}}$)

$\mathrm{\Delta }{\theta }_{\mathrm{MLA}}$ represents the difference between ${\theta }_{\mathrm{MLA},t}$ during the transition from cushioning to push‐off in the stance phase and the initial ${\theta }_{\mathrm{MLA},0}$ at ground contact, as shown in the following equation [7]:

$$\mathrm{\Delta }{\theta }_{\mathrm{MLA}}={\theta }_{\mathrm{MLA},t}-{\theta }_{\mathrm{MLA},0}$$

2.5.3. MLA Moment ($M_{\mathrm{MLA}}$)

$M_{\mathrm{MLA}}$ represents the moment at the apex of the medial longitudinal arch, defined as the net moment acting on the navicular tuberosity (N m/kg). In the equation, the MLA refers to the medial longitudinal arch, COP is the center of pressure, d is the displacement vector from the FNT to the COP, and GRF is the ground reaction force. From the resulting moment vector, two components were extracted: the sagittal‐plane component (around the medial–lateral axis) represents the dorsiflexion/plantarflexion moment, and the frontal‐plane component (around the anterior–posterior axis), representing the inversion/eversion moment acting on the MLA. The specific calculation formula is as follows [7]:

$$M_{\mathrm{MLA}}=\overrightarrow{d}\times \overrightarrow{\mathrm{GRF}}$$

2.5.4. First Metatarsophalangeal Joint (1st MTPJ) Angle

According to the IOR multi‐segment foot model [24], the foot is defined as four segments, comprising the hindfoot, midfoot, forefoot/metatarsus, and hallux. The angle between the hallux and forefoot represents the 1st MTPJ angle. Two components of the 1st MTPJ angle were extracted: the sagittal‐plane component (around the medial–lateral axis), representing the dorsiflexion/plantarflexion angle, and the frontal‐plane component (around the anterior–posterior axis), representing the inversion/eversion angle of the 1st MTPJ.

2.6. Statistical Analysis

Statistical analysis of the eigenvalues was performed using SPSS. The normality of the distribution was assessed using the Shapiro–Wilk test. If the data followed a normal distribution, a mixed‐design two‐way ANOVA was conducted. If an interaction effect existed between the independent variables, a separate analysis of the main effects for each factor was performed. If the data did not follow a normal distribution, the Wilcoxon signed‐rank test (for related samples) or Mann–Whitney U test (for independent samples) was used. Differential analysis of the curves was performed using one‐dimensional statistical parametric mapping (1‐d SPM) in MATLAB (R2024a). The outputs of 1‐d SPM included a time series of t values, which allowed us to analyze differences across the entire stance phase. For normally distributed data, 1‐d SPM was used, and for non‐normally distributed data, one‐dimensional statistical non‐parametric mapping (1‐d SnPM) was applied. The significance level was set at α = 0.05, and all data are presented as the mean ± standard error.

3. Results

3.1. Differences Between NF and FF Under Different Gait Loads

3.1.1. MLA Angle

The study found an interaction effect between gait load and foot type on the MLA angle compression (p = 0.003). Both gait load and foot type had main effects on the peak MLA angle (p = 0.018, p = 0.001, respectively) (Table 2). Regardless of whether walking or running, the MLA angle throughout the stance phase was greater in the FF group than in the NF group (Walking: p = 0.001, Running: p = 0.001) (Figure 3A,B). In the NF group, between 20% and 55% of the stance phase, the MLA angle during running was higher than that during walking (p = 0.001), whereas between 74% and 101% of the stance phase, the MLA angle during running was lower (p = 0.001) (Figure 4A).

TABLE 2.

Mean and standard error of compression of the MLA angle, MLA angle, MLA moment and 1st MTPJ angle in NF and FF groups under different gait loads.

Variables	Walking		Running		p (partial η 2)
	NF	FF	NF	FF	Foot type	Gait load	Interaction effect
MLA angle compression (°)	3.85 ± 1.84	4.42 ± 2.12	4.88 ± 1.99 a	3.27 ± 2.52 b	0.150 (0.015)	0.856 (0.000)	0.003 (0.063)
MLA peak angle (°)	142.68 ± 6.04	149.08 ± 5.47 b	144.09 ± 6.50 a	152.69 ± 6.02 a , b	0.001 (0.274)	0.018 (0.041)	0.296 (0.008)
1st MTPJ peak dorsiflexion angle (°)	93.97 ± 4.96	92.38 ± 4.18	90.42 ± 4.48 a	91.15 ± 3.80	0.586 (0.002)	0.003 (0.065)	0.146 (0.016)
1st MTPJ minimum dorsiflexion angle (°)	64.92 ± 0.71	60.07 ± 0.66 b	62.58 ± 0.65 a	51.15 ± 0.84 a , b	0.001 (0.266)	0.001 (0.086)	0.693 (0.001)
1st MTPJ peak inversion angle (°)	−7.33 ± 0.85	−9.66 ± 0.68 b	−6.44 ± 0.93 a	−9.47 ± 0.81 b	0.003 (0.063)	0.538 (0.003)	0.694 (0.001)
1st MTPJ minimum inversion angle (°)	−4.86 ± 0.90	−7.06 ± 0.59	−4.14 ± 0.89	−5.51 ± 0.73 a	0.041 (0.030)	0.192 (0.012)	0.633 (0.002)
MLA peak dorsiflexion moment (N m)	0.43 ± 0.02	0.50 ± 0.04	0.63 ± 0.04 a	0.75 ± 0.10 a	0.064 (0.025)	0.001 (0.120)	0.616 (0.002)
MLA peak plantarflexion moment (N m)	−0.87 ± 0.29	−0.61 ± 0.62 b	−1.47 ± 0.26 a	−1.25 ± 1.04 a	0.018 (0.041)	0.001 (0.219)	0.879 (0.000)
MLA peak eversion moment (N m)	0.69 ± 0.01	0.56 ± 0.02 b	1.54 ± 0.03 a	1.45 ± 0.04 a	0.001 (0.104)	0.001 (0.882)	0.517 (0.003)
MLA peak inversion moment (N m)	NA	−0.26 ± 0.05	NA	−0.27 ± 0.07	NA	NA	NA

Note: MLA inversion moment not observed in NF group; denoted as NA. Bold p‐values indicate a significant main effect of a factor or a significant interaction effect between the two factors, p < 0.05.

^a Compared to walking, a significant difference exists in the running condition, p < 0.05.

^b Compared to NF, a significant difference exists in the FF group, p < 0.05.

FIGURE 3.

Time‐series curves and SPM results of sagittal plane parameters throughout the stance phase (comparison among NF group, FF group, and FF group with insoles). The shaded regions above each subplot indicate time intervals during which the SPM analysis revealed statistically significant differences between the curves.

FIGURE 4.

Time‐series curves and SPM results of sagittal plane parameters throughout the stance phase (comparison between walking and running conditions). The shaded regions above each subplot indicate time intervals during which the SPM analysis revealed statistically significant differences between the curves.

3.1.2. First MTPJ Angle

Both gait load and foot type had main effects on the minimum dorsiflexion angle of the 1st MTPJ (p = 0.001, p = 0.001, respectively) (Table 2). Under both walking and running conditions, the dorsiflexion angle of the 1st MTPJ in the FF group was lower than that in the NF group (Figure 3E,F). Foot type had a main effect on the peak inversion angle of the 1st MTPJ (p = 0.003) (Table 2). Under both walking and running conditions, compared with NF, the 1st MTPJ in FF exhibited a greater inversion angle, but this difference was statistically significant only during 72%–101% of the running stance phase (p = 0.013) (Figure 5C,D).

FIGURE 5.

Time‐series curves and SPM results of frontal plane parameters throughout the stance phase (comparison among NF group, FF group, and FF group with insoles). The shaded regions above each subplot indicate time intervals during which the SPM analysis revealed statistically significant differences between the curves.

3.1.3. MLA Moment

Both gait load and foot type had main effects on the peak plantarflexion moment of the MLA (p = 0.001, p = 0.018, respectively), whereas only gait load had a main effect on the peak dorsiflexion moment of the MLA (p = 0.001) (Table 2). During walking, the FF group had a smaller peak plantarflexion moment of the MLA compared to the NF group (p = 0.001). During running, the FF group reached the peak dorsiflexion and plantarflexion moments of the MLA earlier than the NF group (Figure 3C,D). Both gait load and foot type had main effects on the peak eversion moment of the MLA (p = 0.001, p = 0.001, respectively). During walking, the FF group exhibited a smaller peak eversion moment of the MLA compared to the NF group (p = 0.001) (Table 3). Observing the MLA moment curve in the coronal plane, we found that during 0%–7% of the stance phase while walking and 0%–13% while running (p = 0.001, p = 0.001, respectively), the NF group exhibited an eversion moment, while the FF group initially showed an inversion moment, which then shifted to a smaller eversion moment compared to the NF group (Figure 5A,B). Moreover, compared with walking, the NF group exhibited a greater MLA eversion moment during 0%–74% of the stance phase (p = 0.001) and a lower moment during 76%–97% (p = 0.001) while running. Similarly, the FF group demonstrated a higher MLA eversion moment during 13%–73% of the stance phase (p = 0.001) and a lower moment during 77%–94% (p = 0.001) when running (Figure 6A,B).

TABLE 3.

Mean and standard error of compression of the MLA angle, MLA angle, MLA moment and 1st MTPJ angle in the FF group before and after wearing orthotic insoles under different gait loads.

Variables	Walking		Running		p (partial η 2)
	FF + insole	FF	FF + insole	FF	Insole	Gait load	Interaction effect
MLA angle compression (°)	4.98 ± 2.47	4.42 ± 2.12	5.37 ± 2.29 c	3.27 ± 2.52	0.003 (0.077)	0.387 (0.007)	0.087 (0.027)
MLA peak angle (°)	144.19 ± 5.23 c	149.08 ± 5.47	146.50 ± 5.89 a , c	152.69 ± 6.02 a	0.001 (0.199)	0.007 (0.066)	0.544 (0.003)
1st MTPJ peak dorsiflexion angle (°)	93.70 ± 3.72	92.38 ± 4.18	90.23 ± 4.68 a	91.15 ± 3.80	0.805 (0.001)	0.004 (0.078)	0.173 (0.019)
1st MTPJ minimum dorsiflexion angle (°)	61.70 ± 1.70 c	60.07 ± 0.66	60.56 ± 0.92 c	51.15 ± 0.84 a	0.024 (0.046)	0.068 (0.030)	0.422 (0.006)
1st MTPJ peak inversion angle (°)	−11.12 ± 0.62 c	−9.66 ± 0.68	−11.22 ± 0.70 c	−9.47 ± 0.81	0.025 (0.046)	0.946 (0.000)	0.832 (0.000)
1st MTPJ minimum inversion angle (°)	−7.60 ± 0.93 c	−7.06 ± 0.59	−7.85 ± 0.72 c	−5.51 ± 0.73 a	0.058 (0.033)	0.390 (0.007)	0.233 (0.013)
MLA peak dorsiflexion moment (N m)	0.51 ± 0.06	0.50 ± 0.04	0.78 ± 0.12 a	0.75 ± 0.10 a	0.824 (0.000)	0.003 (0.078)	0.918 (0.000)
MLA peak plantarflexion moment (N m)	−0.76 ± 0.14	−0.61 ± 0.62	−1.06 ± 1.16 a	−1.25 ± 1.04 a	0.899 (0.000)	0.004 (0.073)	0.286 (0.011)
MLA peak eversion moment (N m)	0.55 ± 0.05	0.56 ± 0.02	1.36 ± 0.06 a	1.45 ± 0.04 a	0.292 (0.010)	0.001 (0.770)	0.401 (0.007)
MLA peak inversion moment (N m)	−0.37 ± 0.09	−0.26 ± 0.05	−0.84 ± 0.17 c	−0.27 ± 0.07	0.002 (0.087)	0.023 (0.047)	0.033 (0.042)

Note: Bold p‐values indicate a significant main effect of a factor or a significant interaction effect between the two factors, p < 0.05.

^a Compared to walking, significant differences exist in the running condition, p < 0.05.

^c Compared to not wearing insoles, the FF group showed a significant difference after wearing orthotic insoles, p < 0.05.

FIGURE 6.

Time‐series curves and SPM results of frontal plane parameters throughout the stance phase (comparison between walking and running conditions). The shaded regions above each subplot indicate time intervals during which the SPM analysis revealed statistically significant differences between the curves.

3.2. Biomechanical Changes in Flatfoot Runners After Wearing Orthotic Insoles

3.2.1. MLA Angle

The orthotic insole factor had a main effect on the compression of the MLA angle (p = 0.003), and both orthotic insoles and gait load had main effects on the peak MLA angle (p = 0.001, p = 0.007, respectively) (Table 3). When wearing orthotic insoles while running, the MLA angle compression in the FF group significantly increased compared to before wearing the insoles (p = 0.001). Under both walking and running conditions, after wearing orthotic insoles, the MLA dorsiflexion angle of flatfoot runners significantly decreased throughout the entire stance phase (p = 0.001, p = 0.001, respectively), aligning with the level of the NF group (Figure 3A,B).

3.2.2. First MTPJ Angle

The orthotic insole factor had a significant main effect on the minimum dorsiflexion angle of the 1st MTPJ (p = 0.024) (Table 3). Under both gait loads, wearing orthotic insoles increased the dorsiflexion of the 1st MTPJ in flatfoot runners, bringing the curve closer to that of the NF group, yet significant differences remained compared to the NF group during 82%–96% of the walking stance phase (p = 0.002) and 57%–86% of the running stance phase (p = 0.001) (Figure 3E,F). The orthotic insole factor had a significant main effect on the peak inversion angle of the 1st MTPJ (p = 0.025). Under both gait loads, wearing orthotic insoles caused the 1st MTPJ of flatfoot runners to become more inverted (Figure 5C,D).

3.2.3. MLA Moment

The gait load factor had a main effect on the peak plantarflexion moment (p = 0.004) and the peak dorsiflexion moment (p = 0.003) of the MLA (Table 3). Compared to when not wearing insoles, the MLA dorsiflexion moment in the FF group wearing insoles decreased only during the early stance phase of walking (6%–27%, p = 0.001), with no significant differences observed in other phases or during running (Figure 3C,D). The gait load factor had a main effect on the peak eversion moment of the MLA (p = 0.001), while the orthotic insole and gait load factors interacted to affect the peak inversion moment of the MLA (p = 0.033) (Table 3). Observing the MLA moment curve in the coronal plane, we found that during 17%–76% of the stance phase while walking (p = 0.001) and 12%–32% of the stance phase while running (p = 0.001), flatfoot runners showed a significant reduction in MLA eversion moment after wearing orthotic insoles compared to those not wearing them, although the overall trend of the curves for both groups remained unchanged (Figure 5A,B).

4. Discussion

Our results revealed that as the gait load increased, the compression of the MLA in flatfoot reduced, and the propulsion effect during running was similar to that during walking. Orthotic insoles alter the kinematics of the flatfoot MLA through physical support, resulting in a movement pattern that is more similar to that of a normal foot. However, the MLA moment remained unchanged compared with that observed without the insoles. This may be because flatfoot runners immediately adjust their COP and GRF when wearing orthotic insoles to maintain their habitual kinetic pattern.

4.1. Differences Between NF and FF Under Different Gait Loads

4.1.1. MLA Angle

We identified an interaction effect between gait load and foot type on MLA angle compression. In the NF group, MLA compression was greater during running than walking, while in the FF group, no significant difference in MLA compression was observed between running and walking (p = 0.112) (Table 2). This indicates that the cushioning effect of the normal arch is higher during running than during walking, whereas the compression of the flatfoot arch is similar during both running and walking. Therefore, as gait load increases, the compression of the MLA in flatfoot runners decreases, which is consistent with our hypothesis. The restriction of MLA compression increases metabolic cost during running, which is caused by the lower compressive potential of the flatfoot MLA [2]. Additionally, studies have shown that a high, stiff MLA also reduces arch compression capacity [27], suggesting that gait propulsion efficiency may have a nonlinear relationship with MLA compressibility and may not depend solely on MLA stiffness.

Regardless of whether walking or running, the MLA angle throughout the stance phase was greater in the FF group than in the NF group (Walking: p = 0.001, Running: p = 0.001) (Figure 3A,B). This is consistent with previous studies and is caused by the characteristic flattening of the medial longitudinal arch in flatfoot runners [7]. In the NF group, the MLA angle during running was higher than during walking in the mid‐stance phase, but lower than during walking in the early swing phase (Figure 4A). This indicates that, compared to walking, a normal MLA requires stronger cushioning ability in the mid‐stance phase during running, and shows greater rebound amplitude and stronger push‐off ability in the early swing phase. The difference in the MLA angle between running and walking in flatfoot only exists during the gait cushioning phase, with no differences observed during the push‐off phase (Figure 4B). This situation arises because the MLA angle at initial foot strike during running is approximately 5° higher than during walking, which may be due to the influence of movement load on the spatial movement of the foot bones [28]. Furthermore, flatfoot runners generally have poor midtarsal joint locking [7], so when faced with stronger gait loads, their ability to maintain arch shape is worse (Figure 4B).

4.1.2. First MTPJ Angle

Under both gait loads, the dorsiflexion of the 1st MTPJ in the FF group was lower than in the NF group, but there was no significant difference in the peak dorsiflexion angle during the push‐off phase of the stance (Table 1; Figure 3E,F). This result suggests that due to the anatomical characteristic of midfoot collapse in flatfoot, there may be a more restricted dorsiflexion range during the stance phase, but the foot can maintain a normal arch level during the critical push‐off phase. Compared with NF, the 1st MTPJ in FF exhibited a greater inversion angle (Figure 5C,D). This may be attributable to the increased forefoot abduction and dorsiflexion angles in FF, as well as the dropping of the proximal end of the first metatarsal [29], which consequently results in greater inversion of the 1st MTPJ. Under both walking and running conditions, the FF group reached the minimum dorsiflexion angle and inversion peak angle of the 1st MTPJ earlier than the NF group (Figures 3E,F and 5C,D), indicating that flatfoot runners initiate push–off action earlier to propel the body off the ground. This is because flatfoot runners have limited dorsiflexion, so they adopt the strategy of early dorsiflexion of the 1st MTPJ to complete the push‐off. However, in one study, when faced with limited dorsiflexion, flexible flatfoot children and adolescents chose an over‐dorsiflexion compensation strategy at the 1st MTPJ to complete the push‐off [30], which may be due to differences in age and athletic habits of the subjects, as well as the greater flexibility and range of motion in the joints of children and adolescents compared with adults.

4.1.3. MLA Moment

In both NF and FF groups, the time taken for the MLA moment to reach the plantarflexion peak during running is earlier than during walking, indicating that during running, the MLA needs to complete the transition from cushioning to push‐off in a shorter period, which requires a higher level of muscle power and control in the arch. During walking, the MLA peak plantarflexion moment in the FF group is smaller than that in the NF group, and there is no significant difference during running; however, the FF group reaches the peak earlier during running (Figure 3C,D), suggesting that with increased gait load, flatfoot runners may activate the arch muscles earlier to complete the push‐off. Observing the MLA moment curve in the coronal plane, we found that normal‐foot runners exhibit an eversion moment throughout the entire stance phase, while flatfoot runners initially show an inversion moment in the early stance phase, which then transitions to a smaller eversion moment compared to normal foot runners (Figure 5A,B). This is due to the increased rearfoot eversion angle and arch collapse at initial foot contact in flatfoot runners, causing the body's center of mass to shift inward, prompting a change in muscle activation patterns to maintain stability during movement [31].

4.2. Biomechanical Changes in Flatfoot Runners After Wearing Orthotic Insoles

4.2.1. MLA Angle

Under both walking and running conditions, after wearing orthotic insoles, the MLA dorsiflexion angle of flatfoot runners significantly decreased throughout the entire stance phase (Figure 3A,B). Therefore, wearing orthotic insoles can change the kinematic performance of the MLA in flatfoot runners throughout the entire stance phase, with the movement pattern being no different from that of normal‐foot runners, whether walking or running (Figure 3A,B). This suggests that wearing orthotic insoles can increase the arch mobility by raising the arch, providing better arch support for individuals with flexible flatfoot [16]. This is consistent with the findings of previous studies that reported specific kinematic changes, including increases in static and dynamic arch height and increases in compression and rebound amounts [32, 33, 34, 35, 36].

4.2.2. First MTPJ Angle

A comparison of flatfoot runners before and after wearing orthotic insoles showed that wearing orthotic insoles increased the dorsiflexion of the 1st MTPJ under both gait loads (Figure 3E,F). However, the orthotic insoles used in this study provide only arch support, with a flat design in the forefoot. Therefore, this change is more likely due to the morphological support provided by the insoles, which elevate the MLA, thereby reducing the angle between the hallux and the forefoot, leading to a passive increase in dorsiflexion of the 1st MTPJ. This differs from the active dorsiflexion of the 1st MTPJ, which, according to previous studies, promotes the windlass effect, increases foot stiffness, and thereby improves gait propulsion efficiency [37]. However, recent research has questioned the idea that the windlass mechanism dominantly drives increased arch stiffness and gait efficiency, arguing instead that arch stiffness is regulated by multiple structures, including soft tissues, and that an increase in arch stiffness does not necessarily enhance gait efficiency [2]. Therefore, the increase in dorsiflexion of the 1st MTPJ observed in this study should not be interpreted as an indicator of improved gait efficiency. Moreover, under both walking and running conditions, significant differences in the 1st MTPJ angle in the coronal plane were observed throughout the entire stance phase, with the curve deviating further from that of the NF group (Figure 5C,D). This indicates that wearing orthotic insoles did not improve, and may have even worsened, the excessive inversion of the 1st MTPJ in flatfoot runners. This also highlights the limitations of orthotic insoles that only provide arch support, as they may not be suitable for flatfoot runners with varying degrees of severity [38].

4.2.3. MLA Moment

Compared with the NF group, wearing orthotic insoles did not bring the MLA moment curve of the flatfoot closer to the level of the normal foot (Figure 3C,D), with the same trend observed in the coronal plane (Figure 5A,B). These results are consistent with previous studies [39]. This may be attributed to the body's tendency to maintain moment stability by adjusting the direction of the ground reaction force (GRF) and the position of the center of pressure (COP) when changes in the MLA angle occur. Previous studies have shown that individuals with flatfoot exhibit a more medially shifted COP and a greater medial component of the GRF, which may compensate for the reduced moment arm [37, 40]. Similarly, when individuals with flatfoot engage in activity while wearing orthotic insoles, the body adjusts the direction of GRF and the position of COP, thereby influencing the forces and moments acting on proximal joints [41]. However, the immediate intervention effect only reflects the compensatory potential of the existing movement pattern. Because movement patterns develop as long‐term adaptations to individual structural characteristics, long‐term follow‐up is necessary to determine whether orthotic insoles truly induce structural‐functional remodeling in flatfoot.

This study has several limitations. First, this study recruited only 10 adult male flatfoot rearfoot strikers. Although the sample size met the power analysis requirements, it remains relatively small, and the lack of diversity in sex, age, and running patterns limits the generalizability of the results. Second, the use of infrared motion capture based on skin‐mounted markers may have introduced measurement errors due to soft tissue artifacts during dynamic movement. Finally, only a single type of orthotic insole was tested, and the study focused solely on the immediate biomechanical effects of insole use in flatfoot runners, without long‐term follow‐up or comparisons across different insole designs and parameters. Future research should consider these limitations and explore the use of a dual fluoroscopic imaging system (DFIS), which can dynamically capture bones in vivo [42], to follow up on the gait biomechanics of flatfoot populations with different genders and foot strike patterns under different orthotic interventions.

4.3. Perspective

This study investigated the biomechanical performance of the MLA and the 1st MTPJ in normal foot, flatfoot, and flatfoot with orthotic insoles while walking and running. The results showed that compression of the MLA in flatfoot decreased as gait load increased. After wearing arch‐support orthotic insoles, the kinematics of the flatfoot aligned more closely with those of the normal foot, but the kinetics did not change. This indicates that after wearing arch‐support orthotic insoles, flatfoot runners may adjust their foot movement response strategy to maintain their habitual biomechanical pattern. However, future research should note that the reduced MLA angle compression observed in flatfoot runners during running merely reflects a limited capacity for MLA compression under high gait loads and does not directly indicate an increased risk of injury. In addition, this study examined only the immediate biomechanical effects of arch‐support orthotic insoles on flatfoot, and longitudinal studies are needed to establish whether such insoles can genuinely induce structural–functional remodeling in flatfoot runners.

Funding

This study was supported by the Basic Scientific Research Program for Universities of the Liaoning Provincial Department of Education (grant LJ212410176002).

Ethics Statement

This study was approved by the Institutional Review Board of Shanghai University of Sport (No. 102772024RT011).

Conflicts of Interest

The authors declare no conflicts of interest.

Liu H., Ye D., Yang Y., and Zhang S., “Effects of Orthotic Insoles on Gait Biomechanics in Runners With Flatfoot Under Different Gait Loads,” Scandinavian Journal of Medicine & Science in Sports 36, no. 1 (2026): e70205, 10.1111/sms.70205.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.