Physical Therapy Korea

Lateral instability of the ankle is one of the most prevalent causes of musculoskeletal ankle injuries. It frequently manifests in the direction of inversion, with contributing factors such as mechanoreceptor damage, impaired proprioception, and muscle weakness in the lateral ankle structures [1-3]. The peroneal muscles, the primary muscles responsible for eversion, contribute to dynamic ankle stability through the regulation of eccentric muscle contraction [4,5]. Insufficient control of peroneal muscles can have both direct and indirect effects on ankle stability. In early functional rehabilitation, isometric strengthening exercises have been selected for improvement of ankle stability [6]. To obtain detailed information about the dynamic strength of the ankle muscles, clinical isometric testing was required to perform at various angles. Therefore, it is important to assess isometric peroneal muscle function and train these muscles to ensure ankle stability.

The primary muscles recruited for eversion are the peroneus longus (PL) and peroneus brevis (PB). It has been theorized that the role of the PL and PB in controlling ankle stability vary because of their different attachments. The PL contributes to both plantarflexion (PF) of the first metatarsal and pronation of the subtalar joint [7]. Specifically, the anterior compartment of the PL contributes to eversion and PF of the foot [8]. The PB, attached to the base of the fifth metatarsal, offers better resistance to calcaneal inversion moments than the PL [9,10]. For these reasons, studies suggesting selective activation of PL and PB based on their different roles have been proposed [10,11]. To investigate the potential for selective activation of both the PL and PB muscles, several studies have examined the differences in muscle activation in response to changes in ankle angle. Kernozek et al. [12] found no significant difference in reaction time between the PL and PB, with changes in ankle angle having a minimal impact on the peroneal muscle reaction time. However, a recent study reported a reduced reaction time in the PL in PF compared to the neutral (N) posture [13]. Subsequently, Donnelly et al. [11] supported the possibility of selective activation by demonstrating that muscle activation of both the PL and PB during eversion was higher in the PF than in the N posture, with the PB exhibiting greater activation than the PL. Despite these efforts, due to the shared nerve innervation between the PL and PB and their involvement in similar ankle movements, assertion of researchers still inconclusive. If the selective activation of the PL and PB is possible, then an effective posture for targeting these muscles can be proposed for clinical applications.

Effective targeting of peroneal muscle function during eversion requires several considerations. First, the biceps femoris (BF) muscle affects eversion owing to its attachment between the ischial tuberosity and the tibial lateral head. Due to its anatomical structure, tibial rotation can occur during ankle eversion. Therefore, the influence of the BF should be considered in research related to ankle eversion exercises. Second, body posture can affect muscle activation, because it is influenced by the effects of gravity [14]. In terms of eversion, the side-lying posture offers greater resistance to gravity than sitting or supine postures, likely resulting in higher peroneal muscle activation. In recent studies, peroneal muscle activation was compared during eversion in the side-lying posture; however, the effect of body posture was not investigated [15,16].

This study aimed to investigate muscle activation of the PL, PB and BF during eversion in different ankle postures (N and PF) and body postures (sitting and side-lying). The objective of our study was to determine the most effective ankle eversion posture for activating the PL, considering the influence of the BF. We hypothesized that the PL, PB and BF activation levels would vary among the four postures, with the side-lying posture eliciting the highest muscle activation in the PL.

1. Participants

This study included 30 healthy individuals with no history of lateral sprains in the ankle within the previous 6 months. An Ankle Joint Functional Assessment Tool (AJFAT) was used to evaluate the level of ankle instability. Participants achieving an AJFAT score of 22 or higher were classified as having no observable signs of ankle instability [8,17]. In cases where both feet met these criteria, the test side was selected as the leg used for kicking the ball [18]. Participants who were unable to complete the test due to any illness or pathology that might influence neuromuscular control and those with a history of lower-extremity surgery were excluded [19]. Participants who had experienced more than two ankle sprains and sustained recent lower-extremity injuries within the last 6 months were also excluded. An increased thickness of the fat layer correlates with a diminished amplitude in surface electromyography (EMG) [14]. Therefore, individuals with a body mass index of 30 or higher (obese) were excluded to ensure muscle activation measurements. Table 1 shows the detailed demographics of the participants. Ethical approval for this study was obtained from the Institutional Review Board of Yonsei University Mirae campus (IRB no. 1041849-202309-BM-177-03). All subjects were informed about the procedures and purpose of the study and signed a consent form.

Table 1 . Participant demographics (N = 30).

Variable	Male (n = 17)	Female (n = 13)
Age (y)	23.9 ± 3.1	23.9 ± 3.4
Height (cm)	175.1 ± 7.1	164.1 ± 5.4
Body mass (kg)	73.4 ± 13.2	60.8 ± 8.0
Recruited ankle (n)
Dominant/nondominant	15/2	10/3
Right/left	13/4	10/3
AJFAT score (n)	28.1 ± 4.4	29.2 ± 7.0

Values are presented as mean ± standard deviation or number only. AJFAT; Ankle Joint Functional Assessment Tool..

The sample size was determined using G*Power software (ver. 3.1.9.2; Heinrich Heine University Düsseldorf). The input parameters were set as follows: an effect size of 0.22, α = 0.05, and a power of 0.80. The effect size was calculated using the partial eta squared value obtained from the body-ankle interaction effect (partial η² = 0.047) based on the ankle evertor strength of five participants obtained through a pilot experiment. It was estimated that 30 participants were needed.

2. Procedures

Prior to the main test, the participants warmed up by walking around the laboratory at a self-selected speed for 3 minutes. This study incorporated four types of eversion based on body and ankle postures: sitting-N (sitting with a N ankle), sitting-PF (sitting with a PF ankle), side-lying-N (side-lying with a N ankle), and side-lying-PF (side-lying with a PF ankle). The hip and knee joint angles were set to 90°. N ankle posture was defined as 0° dorsiflexion and 0° eversion, while the PF posture was defined as 50° PF and 0° eversion [11]. The initial ankle postures were confirmed using a standard goniometer. During eversion, participants maintained their toes in a N position without flexion or extension. The maximal isometric strength of eversion was measured simultaneously with muscle activation. The order of the four postures was randomized, and maximal eversion was performed twice, each time for 5 seconds. Submaximal strength was determined as 70% of the lowest recorded maximal strength value among the four maximal eversion attempts. The participants were provided with a tablet that displayed the strength sensor values in real time, and they repeated the four eversion postures twice for 5 seconds each, targeting submaximal strength. Similarly, the order of measurements was randomized, and muscle activation and submaximal strength data were simultaneously collected. To mitigate the potential effects of muscle fatigue, there was a 10-second intertrial interval, and a 1-minute break was provided between changes in posture. For the analysis of both strength and muscle activation, data were collected in the middle 3 seconds of each trial, and the mean of the two trials was calculated.

3. Strength Measurements

A Smart KEMA strength sensor (KOREATECH, Inc.) was used to measure the strength (in kgf units) of the ankle evertor muscles. The strength sensor displayed tension when pulled from both ends and had good to high intra-rater reliability (ICC_3,1 > 0.85, ICC_2,1 > 0.85) [20,21]. Each end of the strength sensor was connected to a strap and belt using a knock-type hook. A 5-cm-wide strap was fastened to the distal end of the metatarsal bone of the participants. An adjustable nonelastic belt was fixed to the floor using an adsorber, and its length was adjusted to create an initial tension of 2 kgf on the strength sensor. For the sitting posture, the participants were instructed to sit on a table (Figure 1A). The height of the table was adjusted to ensure that the hip and knee joints were at an angle of 90°. To measure eversion in PF posture, a half-foam roller was placed under the foot with 50° of PF, while the examiner adjusted the height of the table to maintain 90° at the hip and knee joints (Figure 1B). For the side-lying posture, the participants were instructed to flex the hips and knees, allowing half of the foot to be off the table with the knee supported by a towel (Figure 2). The ankle of the upper leg was involved, and the hip and knee joints were maintained at 90° angles. Eversion strength was normalized to body mass.

Figure 1. Electrode attachment sites (A) and settings for eversion in a sitting posture (B).

Figure 2. Participant set up for eversion in a side-lying position and tablet displaying real-time tension of pulling sensor.

4. Surface Electromyography

EMG data were collected using a Tele-Myo DTS equipped with a wireless telemetry system (Noraxon Inc.) at a sampling rate of 1,000 Hz. Data analysis was carried out using MyoResearch XP Master Edition software (Noraxon Inc.). Prior to electrode placement, the skin was shaved and cleaned. Two separate bipolar (Ag/AgCl) surface electrodes were placed on the PL, PB and BF muscles parallel to the fibers of the muscle bellies. The distance between the electrodes was 2 cm. For PL and PB, electrodes were attached distal to the fibular head at one-fourth and three-fourths of the fibular length [11,22]. To ensure minimal cross talk, the muscle activation recorded during eversion with the ankle in the N and PF conditions was confirmed to be generated by PL and PB [11]. For the BF, electrodes were attached to the lateral aspect of the thigh, two-thirds of the distance between the greater trochanter and the knee joint [14]. Data were filtered with a 10–500 Hz band-pass filter and smoothed using a 150-ms moving window [16,23]. Muscle activation was normalized to the maximal voluntary isometric contraction (MVIC). MVIC was defined as the highest value among the muscle activations obtained from the four maximal eversion trials.

5. Statistical Analysis

IBM SPSS Statistics software (ver. 23.0, IBM Co.) was used for statistical analysis. For analysis, the strength variable was determined as maximal isometric eversion strength divided by body mass. The variable for muscle activation was determined as submaximal isometric eversion activation divided by maximal isometric eversion activation, measured concurrently with maximal evertor strength.

To determine the differences in muscle strength and activation based on posture, a 2 × 2 repeated measures analysis of variance (ANOVA) was conducted. This analysis incorporated the within-group factors of body posture (sitting and side-lying) and ankle posture (N and PF). If a significant interaction effect was found, pairwise post hoc comparisons were carried out to explore specific differences between the variables. The level of significance was set at p < 0.05.

To assess agreement between measurement postures, a Bland-Altman 95% limit of agreement (LOA) test was conducted [24]. Agreement between the measurements was plotted as mean against difference, with a mean difference ± 1.96 standard deviations (SD) representing the LOA between measurement methods.

Activation of all muscles had significant effects on ankle and body postures (Table 2). The PL muscle activation showed a significant interaction effect between ankle and body postures (F_1,29 = 9.94, p < 0.05, partial η² = 0.26) (Figure 3). A post-hoc t-test showed a significant difference between body postures regardless of the ankle angle (p < 0.05), with a more pronounced difference in body posture in the PF compared to the N. There was also a significant difference between ankle postures regardless of body posture (p < 0.05), and the difference in ankle posture was more prominent in the side-lying than in the sitting posture.

Table 2 . Muscle activation during ankle eversion in four postures.

Muscle	Sitting			Side-lying			p-value
	N	PF		N	PF		Body	Ankle	Body × Ankle
Peroneus longus	26.0 ± 13.0	33.0 ± 16.1		33.1 ± 14.6	54.1 ± 19.0		< 0.05	< 0.05	< 0.05
Peroneus brevis	30.8 ± 12.8	41.3 ± 16.5		42.8 ± 17.6	55.0 ± 17.2		< 0.05	< 0.05	> 0.05
Biceps femoris	8.9 ± 7.0	15.0 ± 10.3		16.8 ± 9.4	22.0 ± 10.7		< 0.05	< 0.05	> 0.05

Values are presented as mean ± standard deviation. N, neutral; PF, plantarflexed..

Figure 3. PL, PB, and BF muscle activation during ankle eversion in four postures. PL, peroneus longus; PB, peroneus brevis; BF, biceps femoris; N, neutral; PF, plantarflexed; %MVIC, percent maximal voluntary isometric contraction. *p < 0.05.

There were significant main effects of ankle (F_1,29 = 45.37, p < 0.05, partial η² = 0.61) and body postures (F_1,29 = 20.81, p < 0.05, partial η² = 0.42) on PB muscle activation. Muscle activation was significantly higher in the PF than in the N posture and in the side-lying than the sitting posture. There was no significant interaction effect between the ankle and body postures (Figure 3).

There were also significant main effects of ankle (F_1,29 = 50.78, p < 0.05, partial η² = 0.64) and body postures (F_1,29 = 14.53, p < 0.05, partial η² = 0.33) on BF muscle activation. Muscle activation was significantly higher in the PF than in the N condition and in the side-lying than in the sitting condition. There was no significant interaction effect between the ankle and body postures (Figure 3).

There was significant ankle by body interaction effect on evertor strength (F_1,29 = 7.47, p < 0.05, partial η² = 0.21) (Figure 4). A post-hoc t-test showed that evertor strength was significantly higher in the sitting-N than the side-lying-N (p < 0.05). In the PF condition, there was no significant difference between the sitting and side-lying postures (p > 0.05). Evertor strength was significantly higher in the N than the PF condition regardless of body posture (p < 0.05).

Figure 4. Bland-Altman plots indicating the differences of the PL muscle activation (%MVIC) according to four measurement postures. PL, peroneus longus; %MVIC, percent maximal voluntary isometric contraction; N, neutral; PF, plantarflexed.

Overall, most of the differences between measurements were contained within the upper and lower 95% LOA (Figures 5–8). The mean differences and LOA between ankle postures (sitting-N and sitting-PF, side-lying-N and side-lying-PF) were relatively small and narrow compared to the mean differences and LOA between body postures.

Figure 5. Ankle evertor strength during ankle eversion in four postures. N, neutral; PF, plantarflexed. *p < 0.05.

Figure 6. Bland-Altman plots indicating the differences of the PB muscle activation (%MVIC) according to four measurement postures. PB, peroneus brevis; %MVIC, percent maximal voluntary isometric contraction; N, neutral; PF, plantarflexed.

Figure 7. Bland-Altman plots indicating the differences of the BF muscle activation (%MVIC) according to four measurement postures. BF, biceps femoris; %MVIC, percent maximal voluntary isometric contraction; N, neutral; PF, plantarflexed.

Figure 8. Bland-Altman plots indicating the differences of the maximal evertor strength (%body mass) according to four measurement postures. N, neutral; PF, plantarflexed.

This study found significant differences in the muscle activation of the PL, PB, and BF muscles based on ankle and body posture. The side-lying-PF posture demonstrated the highest muscle activation of the PL. All the muscles showed greater activation in the side-lying and PF conditions than in the sitting and N conditions. However, the ankle strength was greater in the side-lying and N conditions than in the sitting and PF conditions.

Considering both PL and PB, muscle activation was greater in the side-lying posture than in the sitting posture. These results could be influenced by the direction of gravity during eversion. Eversion occurs mainly at the subtalar joint and involves calcaneal movement relative to the talus along the anterior-posterior axis [25]. In the sitting posture, the force is applied perpendicularly to the direction of gravity, whereas in the side-lying posture, the force is applied in a direction opposite to that of gravity. Although isometric eversion does not involve the actual joint movement, the higher muscle activation in the side-lying posture may be attributed to the direction of the applied force relative to that of gravity.

Furthermore, differences in muscle activation were found depending on ankle posture. The results showed higher activation in the PF posture for both the PL and PB muscles. In a previous study [11], it was reported that the peroneal muscle activation was higher in PF compared to N posture during isometric ankle eversion in supine (mean difference of muscle activation between N and PF, PL = 7.8, PB = 4.2). Greater muscle activation is expected in the muscles with the greatest mechanical advantage [26]. Muscle activation can change with variations in joint angles [27,28]. Anatomically, both the PL and PB pass posterior to the lateral malleolus and participate in the eversion/pronation of the subtalar joint and PF of the talocrural joint [7]. Therefore, greater muscle activation is anticipated in the PF posture than in the N posture.

The objective of our study was to identify a posture that effectively activates the PL. Among the postures we assessed, the side-lying-PF posture exhibited the highest PL muscle activation. However, there was little difference in muscle activation between the PL and PB in the side-lying-PF posture. Klein et al. [5] reported that PL and PB have nearly identical moment arms, with PL at 21.8 mm and PB at 20.5 mm. The similarity in anatomical pathways and moment arms may result in similar muscle activation, despite differences in their attachments. Nevertheless, side-lying-PF is recommended as a non-weight-bearing evertor-strengthening posture for PL. Interestingly, although muscle activation was greater in the PF posture, the evertor strength was greater in the N posture (Evertor strength in sitting, N = 15.0 ± 6.1 kgf/kg, PF = 11.6 ± 4.3 kgf/kg; evertor strength in side-lying, N = 19.7 ± 7.5 kgf/kg, PF = 12.7 ± 4.1 kgf/kg). Ahn et al. [15] reported that the side-lying isometric evertor strength was greater than in the N posture compared to PF posture (N = 20.4 ± 3.8 kgf/kg, PF = 12.3 ± 3.0 kgf/kg), it was similar to our results. This may be attributed to the presence of other muscles that contribute to eversion under the N posture. Secondary muscles assist in eversion, although they are not necessarily present in all individuals. Secondary muscles include the peroneus tertius that is attached to the fifth metatarsal bone along with PB, the peroneus quartus that typically originates from PB and is attached to the calcaneus, and peroneus digiti minimi that originates from PB and is attached to the fifth metatarsal bone [29,30]. These secondary muscles share attachments and are closely associated with dorsiflexion eversion. In this study, muscle activation was measured at the same weight of submaximal isometric eversion for all four postures. If secondary muscles are activated concurrently in the N condition, the muscle activation of the PL and PB may not be relatively high.

Furthermore, among the four of eversion exercises investigated, the side-lying-PF posture emerged as the most promising for evertor strengthening. Previous studies have suggested that the effects of strength training can be expected when muscle activation reaches or exceeds 50–60 %MVIC [31,32]. In this study, side-lying-PF was the only posture where PL muscle activation surpassed 50 %MVIC. However, side-lying-PF concurrently showed that the highest activation occurred in the BF muscle. The BF muscle activation was greater in the side-lying posture than in the sitting posture. Unlike sitting and side-lying eversion involves the knees of the participants touching the floor, providing additional support. With a wider base of support, greater muscle activation can be induced in the side-lying posture. Additionally, the BF muscle activation was significantly greater in the PF than in the N posture. We assumed that increased BF muscle activation may be associated with greater external rotation of the tibia. The participants may have had combined eversion with ankle abduction and tibial external rotation during isometric eversion. A previous study reported a tendency for increased inter-joint pressure in the talotibial joint when the ankle everted in the PF compared to the N posture [33]. If the congruity of the talotibial joint is tight, more compensatory motion can occur to achieve the same eversion level. It is possible that the PF condition induced more tibial rotation as a compensatory mechanism, contributing to increased BF muscle activation. Although the muscle activation of the BF in the side-lying-PF posture was the highest (22 %MVIC), it corresponded to a low-moderate activation level [34,35]. Therefore, we suggest utilizing the side-lying-PF posture, even when considering its slight BF activation.

Our study had several limitations. First, it was not possible to confirm whether the recorded signals originated exclusively from the PL and PB muscles. Unlike invasive electromyography, surface electromyography cannot directly contact muscles, and signals from adjacent muscles can be acquired. In particular, signals from the extensor digitorum longus that is located near the peroneal muscles, can be acquired [36]. Although efforts were made to minimize crosstalk by instructing the participants to avoid flexion or extension of their toes during eversion, it was impossible to eliminate this issue completely. Second, it remains unclear whether tibial rotation occurs during eversion. We found that the BF was activated within the range of 8.9–22.0 %MVIC during eversion. However, it is unknown whether this level of activation is sufficient to induce tibial rotation. The primary role of the BF is related to the flexion and extension of the hip and knee joints. Given that all four eversion exercises in this study maintained consistent hip and knee joint angles, it can be inferred that increased BF activation was linked to tibial rotation rather than to other movements. Third, the outcomes of this study were primarily restricted to submaximal eversion. These methods may be insufficient for athletes or individuals who require higher levels of strength training. However, the results of our study are relevant to learning how to effectively use the peroneal muscles in the early stages of rehabilitation or applying isometric exercises for unstable ankles. Future studies are warranted to compare the effects of these exercises in patients with lateral instability of the ankle to determine their clinical usability.

This study showed that the PL, PB, and BF muscles had significant differences in muscle activation depending on the body and ankle postures. Among the evaluated postures, the side-lying-PF posture showed the highest muscle activation. These findings indicate the impact of body and ankle postures on isometric eversion and suggest that the side-lying-PF eversion posture may be preferred to effectively induce muscle activation of the peroneal muscle, even when considering BF muscle activation.

None.

None to declare.

No potential conflicts of interest relevant to this article are reported.

Conceptualization: DL, OK. Data curation: DL. Formal analysis: DL, JK. Investigation: DL, SH. Methodology: DL, OK. Project administration: DL. Resources: DL, SH. Supervision: JK. Visualization: DL. Writing - original draft: DL. Writing - review & editing: JK, OK.