Introduction
Ankle fractures are one of the most common lower limb fractures [
1]. Based on the number of malleoli involved, ankle fractures can be classified into isolated malleolar fractures, bimalleolar fractures and trimalleolar fractures, which occur in around 66, 25 and 7% of all ankle fractures respectively [
2]. The trimalleolar fractures have worse clinical outcomes and an increased incidence of osteoarthrosis [
2], and a notable proportion was unable to return to sports [
3].
Although some studies investigated the differences among subgroups based on the severity of fractures [
1,
4,
5], there were few studies focused on the characteristics of trimalleolar fractures alone [
6]. Most studies used radiographic assessment, physical examination and patient-reported outcome measures (PROM) to evaluate the functional status of the patients post ankle fractures surgeries [
1,
4]. Although these methods are considered to be valid, they are static and subjective, and dynamic, objective and precise evaluations are needed.
Gait is the most common and fundamental method for humans to perform physical activities, and the integrated actions of the nervous, muscular and skeletal systems are reflected in walking [
7]. Gait analysis is a significant way of characterizing functional performance of humans, including temporal-spatial parameters (TSPs), gait kinematics, gait kinetics and musculoskeletal activity [
8,
9]. TSPs are referred to as the vital signs of gait, including walking speed, cadence, step length and so on [
8]. Plantar pressure distribution (PPD) includes many valuable information for evaluating stability and motor control ability of ankle, and it had been estimated in subjects who were at risk of sustaining ankle injuries or ankle instabilities [
10]. The lower limb muscles are the active part of walking, and analyzing muscle activation during functional tasks, such as walking, would be more comprehensively to evaluate the ankle function [
9]. Several studies have investigated lower extremity muscle activation by surface electromyography (sEMG) in patients with chronic ankle instability, ankle sprains and so on [
7,
11]. Biomechanical studies on gait in patients with trimalleolar fractures are still rare, and to the best of our knowledge, this was the first study to exclusively explore the TSPs, PPD and sEMG characteristics of trimalleolar fractures patients.
The primary aim of this study was to investigate the gait biomechanics including TSPs, PPD and sEMG in patients with trimalleolar fractures during walking, and the results were compared with healthy controls. The secondary aim was to evaluate the association between gait parameters and ankle clinical assessments.
Discussion
This study indicated that at an average of 4.5-months post-surgery, patients with trimalleolar fractures showed poor OMAS results, and their injured ankles were swollen, and the passive ROM were decreased. During gait analysis, patients demonstrated abnormal gait compared with healthy controls, and an asymmetrical gait pattern was seen in patients. Compared with healthy controls, the abnormal performance of patients in plantar pressure distribution were concentrated in hindfoot and forefoot, and patients also showed abnormal muscle activity of TA and PL. Furthermore, the passive inversion ROM was highly correlated to symmetry index of step width and walking. This study was the first to indicate remaining detailed gait deficits in patients with trimalleolar fractures. In addition, gait parameters were correlated with clinical outcomes in patients with trimalleolar fractures for the first time.
OMAS are usually used as a reliable and valid outcome measure after an ankle fracture [
18], and based on the total score, ankle function of patients could be divided into four grades: excellent (OMAS:100 to 91 points), good (OMAS:90 to 61 points), fair (OMAS:60 to 31 points) and poor (OMAS:30 to 0 points) [
19]. In this study, the mean OMAS was 56, indicating that patients with trimalleolar fractures reported fair ankle function. According to the study, the disability of running and jumping contributed most to the total score. Several studies also investigated the OMAS of ankle fractures patients, and they showed better results than our study. Oguzhan Tano glu et al. [
20] compared the effect of a 1-stage surgery for the unstable malleolar fracture dislocations with the 2-stage surgery. The two group all included patients with isolated malleolar fractures, bimalleolar fractures and trimalleolar fractures. And the duration of follow-up of the two group was 21.7 and 19.2 months respectively. The mean OMAS was 87.8 for the 1-stage surgery group and 83.2 for the 2-stage surgery. Mareen Braunstein et al. [
21] demonstrated functional outcomes after 1 year of arthroscopically assisted ankle fracture treatment, and they reported a mean OMAS of 85 for trimalleolar fractures. The poor OMAS results for trimalleolar fractures in this study might be mainly attributed to the short length of the postoperative follow-up period. This study investigated the clinical outcomes after 4.5 months postoperatively, while other studies evaluated the long-term (more than 1 year) clinical outcomes.
This study indicated that patients with trimalleolar fractures remained physical impairments postoperatively. Compared with the noninjured side, patients represented ankle swelling, and a decrease in passive ROM on the injured side. Ankle swelling is a common and long-standing complication after surgery. It was reported that more than half of the patients following unimalleolar and bimalleolar ankle fractures presented stiffness, swelling and pain [
1]. 60% or more of the patients 65 years or older reported ankle pain, swelling and problems when using stairs and reduced activities of daily life 1 year after ankle fractures [
22]. Shah et al .[
23] demonstrated that around 45% of 69 patients with Weber B and C ankle fractures still had ankle swelling at 5 years after the injury. Our study investigated the passive ROM, and mean (sd) angle of dorsiflexion, plantarflexion, inversion and eversion of the injured leg on the trimalleolar fractures patients was 7.08 (3.91) degrees, 33.33 (8.07) degrees, 13.75 (6.78) degrees and 6.00 (4.17) degrees respectively. Ganit Segal et al. [
1] also measured the passive ankle ROM of patients with trimalleolar fractures in the sagittal (dorsiflexion/ plantar flexion) and coronal plane (inversion/eversion), and the ROM was − 0.8 (7.6) degrees, 40.6 (7.5) degrees, 5.6 (3.6) degrees and 2.5 (4.0) degrees respectively. The result was a little different from our study, might due to the assessment point. Ganit Segal et al. [
1] measured ROM at 64.5 days from injury, which was earlier than ours (4.5 months). There was also a study evaluated ROM during activities. Van Hoeve, S et al. [
4] found that compared with the healthy subjects (12.59 ± 3.73 ), the ROM during gait in patients (7.13 ± 2.55 ) with trimalleolar ankle fractures decreased significantly. After ankle fractures, the uncoagulated hemorrhage leads to the rapid increase of intra-articular pressure, which causes abrupt joint swelling, pain and limited mobility [
24]. Presence of soft tissue damages such as tendon and ligamental injuries can cause chronic swelling and stiffness therefore resulting in the dismal outcome [
3]. These complications might alter the gait.
The present study showed that patients with trimalleolar fractures presented compromised gait pattern. The temporal-spatial parameters of the injured side and noninjured side were significantly different from healthy subjects, and the two sides were also different, except no difference exist in step width. This is in line with other research findings. Three studies investigated the gait parameters of patients with trimalleolar fractures, and compared those with healthy group. They found all gait parameters were significantly below the normal range [
1,
4,
6]. Andrew F. Tyler et al. [
6] also showed that the gait characteristics of ankle fracture patients were more similar to healthy elderly patients. However, these studies only investigated the differences between patients and healthy subjects, but did not compare the gait parameters of the injured side with the noninjured side. Ganit Segal et al. [
1] also examined limb symmetry of the gait patterns, and found significant asymmetry in step length and single limb support. These findings were consistent with this study. And by correlation analysis in this study, step asymmetry might be related to the difference of ankle inversion ROM between two sides. All these results presented that patients with trimalleolar fractures did not achieve restoration of normal physiologic gait in the short-term, and patients adopted a simple security strategy with a reduction of walking speed [
7]. In addition, many studies indicate that gait changes over an individual’s lifetime [
25,
26]. With aging, there are physiological changes in the sensorimotor systems, which may cause gait abnormalities. In this study, the age span for each group is relatively large, this may be a bias factor contributing to the abnormal gait features. The results should be further confirmed by shrinking the age span.
The differences of features of plantar pressure distributions and sEMG in both sides also indicated asymmetries in gait in patients with trimalleolar fractures. The plantar pressure in the T345 and the contact area of MF of the injured side were significantly smaller than those of the noninjured side, and this was probably due to a more cautious and compensatory walking pattern, by further biasing the center of gravity to the noninjured side. Sjoerd Kolk et al .[
27] also showed subtle asymmetries in gait kinetics and kinematics between the operated and non-operated limbs, and they considered that patients performed a more cautious walking pattern and an integral strategy. Plantar pressures of other type ankle fractures such as pilon fractures, calcaneal fractures were also asymmetry, and adhesion or conduction disorders at the tibia may be causes of abnormal plantar pressure [
28,
29].
Compared with normal healthy subjects, patients with trimalleolar fracture performed abnormal gait during walking support period. Patients tended to step more cautiously on the injured heel, showing smaller peak plantar pressure in HF, and this might be due to pain or psychological factors such as fear or worry of reinjury [
29]. For patients, the contact time (%) of HF and MF and total contact time were significantly increased, and it might be associated with lower ankle stability: for patients with trimalleolar fractures, the lateral, medial and posterior malleolus were injured, probably impairing the ankle stability, and patients needed more time to maintain ankle stability [
1]. The muscle activity on the TA and PL of the injured side was significantly larger than those in the noninjured, also indicating the ankle stability of the injured side decreased, because muscular co-contraction of TA and PL was increased to stabilize the ankle joint [
30]. And in this study, the step width was significantly smaller than that of healthy controls, also showing the walk stability decreased in patients. The plantar pressure features (smaller peak plantar pressure and contact area, shorter contact time) in the forefoot demonstrated that the propulsion ability during walking significantly decreased in patient, and this could decrease the walking speed. In this study, the IEMG of PL of patients was significantly increased, indicating the muscle ability decreased. Therefore, it was reasonable to speculate that the abnormal gait features in the forefoot might be related to the decreased ability of PL. During normal walking, the plantar pressure in the first metatarsal head plays a key role in pushing off [
31]. And the PL is essential for maintaining ankle stability and plays an important role in the push-off stage. The PL origins at the proximal tibia and fibula, and inserts at the first metatarsal and medial cuneiform [
32]. Except for contributing to 63% of eversion strength, the PL is helpful to initiate pronation and stabilize first ray during propulsion phases of gait [
32].
Many studies indicated that speed has significantly influence on gait kinematics, kinetics, and sEMG. Sander van Hoeve et al. [
4] found that when asked to walk at preferred normal speed, patients after ankle fractures showed lower walking speed compared to healthy subjects. When healthy subjects walked at equal speed, significantly lower ROM in the ankle flexion/extension during loading and push-off phases was also found in patients. And the smallest ankle flexion/extension ROM was found for the patients with trimalleolar fractures. Rachel M. Koldenhoven et al. [
33] analyzed differences in gait characteristics between individuals with chronic ankle instability and healthy controls at three different walking speed (Preferred walking speed (PWS), 120% PWS, and standardized walking speed of 1.34 m/s). It found that the increases in walking speed magnified the differences between the groups for ankle inversion and hip adduction kinematics. But no group differences were identified for EMG variables, which might be due to the relatively small sample size and naturally high variability when measuring EMG activity. Gordon L. Warren [
34] assessed the effects of walking speed (walked on a treadmill at seven speeds between 0.45 and 1.79 m/) on plantar pressures and lower-leg muscle activities in healthy participants. Except for the medial midfoot (
P = 0.33), there were significant effects of speed on the peak values in the pressure–time curves for all plantar regions. There was also a highly significant effect of speed on the peak values in the root mean square –time curves for both the TA and GM muscles (
P < 0.001). Preferred walking speed [
33] is an individualized component of gait and should be considered when analyzing gait in a laboratory setting. However, standardizing walking speeds when measuring differences between groups is also important, which could improve the reproducibility of results between studies. Therefore, it is reasonable to speculate that walking speed may also affect gait parameters between trimalleolar fractures patients and healthy controls. More work needs to be done in this area to better assess gait performance in patients with trimalleolar fractures.
Limitations
There were some limitations in this study. First, no age-matched control group was tested. The current healthy control group trended toward younger age than the fracture patients. This may be a bias factor contributing to the features noted in the gait analysis and may not be attributable to trimalleolar fractures. And the relatively small sample size may not be representative of the population of trimalleolar ankle fractures. Future work with a larger sample size and smaller age range will be required to fully investigate these characteristics. Second, this study did not divide the patients into “unfixed posterior malleoli group” and “fixed Posterior malleoli group”. However, it was demonstrated that fixation of the posterior malleolus particularly did not appear to improve gait characteristics. Therefore, the patients in this study were not further grouped into two subgroups. Third, the participants in this study were asked to walk at their natural speed, and these values in healthy subjects were found to be faster when compared to the patients, which may affect the comparisons between the other gait biomechanics parameters. Future research should examine the difference in gait patterns when patients walk at the same speed.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.