Ultrasonography of Gluteal and Fibularis Muscles During Exercises in Individuals With a History of Lateral Ankle Sprain Open Access

We conducted a descriptive laboratory study using a cross-sectional design. Our independent variable was group (healthy, coper, LAS, and CAI), and our dependent variables were ARs for each muscle (gluteus maximus [GMax], gluteus medius [GMed], and fibularis muscle group) and resistance condition for tabletop exercises (nonresisted and resisted) and during lateral band walking (lower leg and forefoot band placement).

A convenience sample of 67 individuals (27 males, 40 females; healthy = 16, coper = 17, LAS = 15, CAI = 19) with or without a history of an LAS volunteered to participate in this study (Table 1). This study was part of a larger investigation of foot impairments and functional limitations after LAS and CAI.²¹  All participants completed the Godin Leisure Time Questionnaire and reported being physically active 3 times per week for at least 20 minutes. Volunteers were excluded if they had a self-reported history of fracture or surgery to the leg or foot, had any neurologic disorder that might affect balance (eg, diabetes mellitus, lumbosacral radiculopathy), had any soft tissue disorder (eg, Marfan syndrome, Ehlers-Danlos syndrome), were pregnant, or had any contraindication to manual therapy. Healthy individuals were excluded if they had any history of ankle or foot sprain or pain. All participants provided informed consent, and this study was approved by the Institutional Review Board for Health Sciences Research at the University of Virginia in compliance with all applicable federal regulations governing the protection of human participants (#18550).

Copers were defined as having experienced 1 significant LAS at least 12 months before study participation and no episodes of giving way or perceived instability (Identification of Functional Ankle Instability [IdFAI] score ≤10, Foot and Ankle Ability Measure [FAAM]–Activities of Daily Living [ADL] score ≥99%, and FAAM–Sport score ≥97%).⁴  The LAS group was defined as having sustained a significant LAS that affected function between 2 and 8 weeks before the study. Individuals in the LAS group could have a history of more than 1 ankle sprain and were not included based on FAAM–ADL, FAAM–Sport, or IdFAI score; however, they completed the questionnaires for demographic purposes. Chronic ankle instability participants were classified according to the International Ankle Consortium guidelines.⁶  Individuals with CAI had to have incurred at least 1 significant LAS at least 12 months before participation, had not sprained their ankle within 8 weeks of participation, experienced sensations of giving way or residual perceived instability, and scored >10 on the IdFAI, <90% on the FAAM–ADL, and <85% on the FAAM–Sport. Participants' demographic information is detailed in Table 1.

An ultrasound unit (model Acuson Freestyle; Siemens, Mountain View, CA) with an 8-MHz wireless linear transducer was used to visualize and obtain images of the GMax, GMed, and fibularis muscles. Images were obtained with participants in side-lying position for the tabletop scans and while standing for the lateral band-walking exercises (Figure 1). For the standing and band-walking images of the gluteal muscles, the ultrasound transducer was secured in a foam block and placed on the participant's hip using an elastic band (Figure 1).²²  The transducer was postioned so that both gluteal muscles were visualized in the same image and both fibularis muscles were visualized in the same display (Figure 2).

Heavy resistance was provided for band-walking exercises using green exercise tubing (TheraBand, Akron, OH).

For each exercise, only the affected limb was assessed. In participants with bilateral CAI, the self-perceived worse ankle was tested. The healthy group's limbs were assessed in random order using a Latin square.

Participants performed tabletop exercises while in side-lying position. The CSA images for the fibularis muscles were taken at 50% of the distance between the fibular head and lateral malleolus (Figure 1).¹⁸  The test limb was placed on a support block, and the contralateral limb was bent under the participant. For the resting images, the participant was instructed to relax fully. Nonresistance fibularis images were taken when the individual maximally everted the foot. Resistance images were taken in the same position with manual resistance applied by a researcher.

For the GMax and GMed thickness images, the ultrasound probe was placed midway between the iliac crest and greater trochanter where both muscles could be visualized using a technique similar to one previously described (Figure 1).^22,23  Resting measures were taken with the participant side lying and the legs fully relaxed. Nonresistance hip-abduction images were taken when the participant abducted the hip so that the foot was 25.4 cm above the tabletop. Resistance images were taken in the same position as the nonresistance images; however, manual resistance was provided by a researcher and was not standardized. Participants were asked to provide maximal force against resistance and not let the researcher “break” their position. Contractions for all side-lying tabletop measures were isometric. Three images of each condition (rested, nonresistance, resistance) were obtained for the gluteal and fibularis muscles. All images were captured by a single researcher (J.J.F.) and measured by another researcher (R.M.K.).

The ultrasound transducer was placed in a foam block and secured to the hip of the affected limb using an elastic band in the standing position to visualize the GMax and GMed using a technique similar to that used previously (Figure 1).²²  Participants performed 3 sets of 5 lateral band walks while wearing a heavy resistance–level band (placed in each location: lower leg, forefoot) in the direction of the affected side. The tension on the band was not standardized; however, participants were asked to step as widely as possible against the resistance band. One image for each set of lateral band walks was obtained, for a total of 3 images. The band was placed over the lateral malleolus for the lower leg site and over the metatarsals for the forefoot site. The test limbs of healthy participants were randomly assigned and, therefore, the direction in which to perform the task was assigned based on which limb was included in the analysis. Participants were instructed to maintain a partial-squat position and to take 5 lateral steps as widely as they could in the direction of the test limb. Images were taken when the foot contacted the ground at the maximal width of the step. Our method of band walking was intended to simulate exercises used in clinical rehabilitation.

All ultrasound images were taken by a board-certified orthopaedic physical therapist with 15 years of clinical experience and 2 years of USI experience. The images were measured by an athletic trainer with 3 years of clinical experience and 1 year of USI experience who was blinded to group assignments.

Images were measured using ImageJ (version 1.50i; National Institutes of Health, Bethesda, MD). In each condition, 3 images were obtained. The average of the 3 measurements was used in the calculations for ARs in each position. Muscle-thickness measures (cm) were obtained by measuring from the inferior aspect of the superior border of the muscle to the superior aspect of the inferior border of the muscle at the widest portion (Figure 2).²²  Cross-sectional area was determined by tracing the interior aspect of the muscle border to calculate the area (cm²).¹⁸  Activation ratios for muscle thickness, CSA, and functional ARs were calculated using the following formulas:

We performed separate 4 × 2 repeated-measures analyses of variance for each dependent variable. The between-subjects factor was group (healthy, coper, LAS, CAI) and the within-subject factor was resistance condition (tabletop: nonresistance or resistance; band walk: lowerleg or forefoot resistance). The post hoc Fisher least significant difference test was conducted when a significant main effect or interaction was identified. The a priori level of significance was set at P ≤ .05 for all analyses. All data were analyzed using SPSS (version 24.0; IBM Corp, Armonk, NY).

Intraclass correlation coefficient (ICC [3,1]) values were calculated to determine the intratester reliability of USI measures. Excellent intraclass correlation values were found for the 3 images in the fibularis group (>0.99 for rest, nonresisted, resisted), the GMax (tabletop: >0.99 for rest, nonresisted, resisted; band walking: 0.98 for ankle and forefoot sites), and the GMed (tabletop: >0.99 for rest, nonresisted, resisted; band walking: 0.98 for the ankle, 0.97 for the forefoot).

The groups were not different in age, height, or mass. Differences were present in the number of ankle sprains, months since the last ankle sprain, and IdFAI, FAAM–ADL, and FAAM–Sport scores (P < .001 for all), as expected based on the inclusion and exclusion criteria. Resting measures of thickness and CSA were not different among groups (Table 2).

Significant main effects for resistance condition were observed in the fibularis muscles (P < .001) and the GMax (P = .029) for the tabletop exercises. All groups demonstrated increases in AR for the fibularis muscles with resisted compared with nonresisted eversion (% increase: healthy = 3.7%, coper = 3.9%, LAS = 3.2%, CAI = 4.1%). Both the LAS (6.4%) and CAI (7.2%) groups demonstrated increased ARs in the GMax during hip abduction with resistance versus no resistance. A significant resistance-by-group interaction (P = .007) was identified for the GMed (Table 3). Post hoc results for the significant resistance main effects and resistance-by-group interactions are detailed in Figure 3. No other significant group main effects or interactions for any muscle were observed.

Significant main effects for band position were noted for the GMax (P < .001) and GMed (P < .001) muscles during lateral band walking, and ARs increased when the band was moved from the lower leg to the forefoot. Both the healthy and LAS groups displayed an increased GMax AR in the forefoot band position (4.8% and 8.1% increases, respectively). All groups demonstrated an increased GMed AR with the band in the forefoot position (% increase: healthy = 5.8%, coper = 3.0%, LAS = 4.0%, CAI = 4.3%). Post hoc results for the significant band position main effects are shown in Figure 3. No significant group main effects or interactions for either muscle were present. Activation of the gluteal muscles in the tabletop and band-walking conditions is illustrated in Figure 4.

This study was the first to measure the motor activation of the gluteal and fibularis muscles using USI during functional activities among individuals with or without a history of LAS. Our key findings were the increased activity of the GMax and fibularis muscles during tabletop exercises and both gluteal muscles during band-walking exercises when the tasks increased in difficulty. We saw an increase in AR during resisted motion compared with nonresisted motion in both the GMax and fibularis across groups. During nonresisted and resisted tabletop hip-abduction exercises, the coper group had a lower level of GMed motor activation compared with the healthy, LAS, and CAI groups.

No group differences were demonstrated for fibularis muscle activation, which was somewhat surprising. All groups displayed increased fibularis CSA and AR with resisted versus nonresisted eversion. Our results contradict those of a similar USI study conducted by Lobo et al,¹⁸  who found less resting fibularis longus CSA in individuals with a history of LAS compared with a healthy control group. We were not able to clearly differentiate between the fibularis longus and brevis muscles, so we could not replicate the measurement techniques of Lobo et al.¹⁸  Methodologic differences in how the fibularis muscles were measured during our exercise condition may explain the disparate results between studies. Lobo et al¹⁸  assessed the fibularis longus and brevis as discrete muscles, whereas we measured the 2 muscles as a single functional group.

Our findings were consistent with another study⁷  that involved using magnetic resonance imaging to measure lower extremity muscle volumes and strength in individuals with or without CAI. Although the CAI group presented with decreased eversion strength compared with healthy participants, fibularis longus muscle volume did not differ. The researchers attributed their results to altered neuromuscular function as opposed to the size of the muscle,⁷  which may explain why we observed no differences between groups for the fibularis AR.

Side-lying hip abduction exercises are commonly used in rehabilitation to target gluteal muscle activation.²⁴  Moderate to high GMed EMG muscle activity is needed to carry out this exercise, as demonstrated by prior investigators.²⁴  In our study, participants performed a side-lying hip-abduction exercise. Both LAS and CAI groups showed increased GMax muscle activation with manual resistance, but the coper and healthy groups did not. The central motor alteration that occurs after LAS²⁵  and CAI^26,27  may affect motor recruitment at the hip and explain our results in the tabletop exercises. The possibility of altered hip recruitment is further supported by the earlier findings from our laboratory¹⁰  in which alteration of the GMed during the preswing phase of gait was observed in participants with CAI.

The GMed activation increased from the nonresisted to the resisted condition in the healthy, coper, and LAS groups; however, activation in the CAI group did not change between conditions. In addition, the CAI group had a lower GMed muscle AR with resistance compared with all other groups. Yet when we looked at the gluteal muscle group as a whole, the resisted condition provoked an increase in muscle activation in all groups for at least 1 gluteal muscle. Clearly, the GMed in individuals with CAI functioned differently than in the other groups during side-lying abduction exercises. Individuals with CAI may compensate by using more GMax activation to accomplish the task.

Band-walking exercises have been used to target gluteal muscle activation.²⁴  When we placed bands around the forefoot, the activity for both gluteal muscles increased compared with the lower leg band placement. Thus, extending the length of the lever arm and introducing an externally applied medial-rotation moment by moving the band placement distally increased the difficulty of the task. The GMed activity increased in all groups when the band was placed around the forefoot compared with the lower leg. For the GMax, muscle activity increased from the lower leg to the forefoot band position in the healthy and LAS groups but not in the coper or CAI groups. This may suggest that individuals who were copers or had CAI relied on the GMed more than the GMax to accomplish the band-walking task. Adjusting the position of the band may be beneficial when the goal of rehabilitation is to increase gluteal muscle activity. Increasing the difficulty of the exercise requires more activation from the gluteal muscle groups as a whole, whether the exercise occurs in an unloaded or loaded joint position. Ultrasound imaging could be used to provide visual biofeedback when the rehabilitation goal is to increase the activation of a specific muscle.

No differences in gluteal muscle activation were identified between groups for the band-walking exercises in either band position. This was somewhat surprising as individuals with a history of LAS have demonstrated differences in hip-muscle function during various tasks when compared with healthy control participants.^10,14,28,29  Although the difference was not significant, the coper group used 7% to 12% more GMed muscle activity than all other groups during the band-walking tasks. Increasing GMed activation may be a strategy that copers use to reduce the risk of subsequent LAS. Targeting increased GMed muscle activity in those with CAI and those who have recently experienced an LAS appears to be an appropriate rehabilitation strategy.

Gluteal muscle activation increased similarly during the tabletop and band-walking exercises. Using USI as a platform for visual biofeedback has potential in several areas, including helping patients target activation of specific muscles during rehabilitation. For example, our participants with CAI did not increase GMed activation and relied more on GMax activation during resisted side-lying hip abduction. The USI could allow a patient to visualize the muscle and alter movement according to the goal set by the clinician. Additionally, this tool could be used throughout a rehabilitation program to track muscle activation and hypertrophic changes over time.

Patient positioning was slightly different in our study than in previous research^18,30  to ensure that we could image the fibularis muscles in nonresisted and resisted conditions. Participants performed only uniplanar tasks and, therefore, this information can be applied only to exercises performed in the frontal plane. Lastly, we studied a convenience sample of participants as part of a larger investigation.

As task difficulty increased with added resistance, regardless of the type of exercise, gluteal and fibularis muscle activity also increased. In the side-lying position, individuals with CAI demonstrated no change in GMed activity between resistance conditions and may rely more on GMax activity. During band walking, activity of the gluteal muscles increased when the bands were placed in a more distal position. Furthermore, copers used 7% to 12% more gluteal activation during resisted band-walking tasks, which may suggest an altered hip strategy among individuals with a history of a single LAS. Clinicians may use this information to target hip-muscle activity for patients during rehabilitation. In individuals with a history of LAS, ultrasound imaging could supply visual biofeedback to target more specific muscle activations while performing the prescribed exercises.

We thank the University of Virginia's Curry School of Education Foundation for its generosity in providing funding.

This study was funded in part by the University of Virginia's Curry School of Education Foundation. Neither the Department of the Navy nor any other component of the Department of Defense has approved, endorsed, or authorized this manuscript.