Fractures of the ankle are common. In the United Kingdom the annual incidence of ankle, tibial, and fibular fractures is 14.8 per 10,000 people.17 Ankle fractures are usually treated with 3 to 6 weeks of cast immobilization, sometimes after surgical fixation.10 This immobilization causes reduction in strength18 and ankle flexibility.4,13,20 Pain13 and swelling1 are also common sequelae. These impairments can restrict activities such as walking and ascending and descending stairs.13
Ankle flexibility is altered by the initial ankle fracture injury22 and subsequently by changes associated with immobilization of the soft tissues.2,7,14-16,21,25 Animal studies indicate when joints are immobilized for days or weeks muscle fibers and tendons become short8,16 and periarticular connective tissues become inextensible.2
Four studies have investigated passive ankle flexibility after cast immobilization for ankle fracture.4,5,13,20 They show plantarflexion-dorsiflexion range of motion is decreased,4,5,13,20 stiffness into dorsiflexion is increased,4,20 and hysteresis is increased.4 While two of the studies employed small samples of convenience,4,20 one was a large randomised controlled trial13 and one was a large prognostic study.5 Our clinical trial compared the efficacy of short-duration (ie, 6 minutes/day) and long-duration (ie, 30 minutes/day) passive stretches with a control treatment (ie, exercise only) for the management of plantarflexion contracture after cast immobilization for ankle fracture.13 The addition of passive stretching conferred no benefit over exercise alone in terms of activity limitation and passive dorsiflexion range of motion. Initial orthopaedic management (surgical versus non-surgical) was one predictor of outcome after cast removal for ankle fracture.5
The primary aim of this investigation was to quantify the recovery of ankle flexibility in people with ankle fracture treated with cast immobilization. We hypothesized recovery would be incomplete 3 months after cast removal. The secondary aim was to determine if initial orthopaedic management (ie, surgical or nonsurgical) was a predictor of the recovery of ankle flexibility. We hypothesized people with nonsurgical management would have better ankle flexibility than those who required surgical fixation. To attain these aims we performed additional analyses, including three previously unreported flexibility variables, on data collected in our randomized controlled trial of passive stretching.13
MATERIALS AND METHODS
We recruited people with ankle fractures treated with cast immobilization, with or without surgical fixation, from the fracture clinics of two large teaching hospitals. Inclusion criteria were: referral to the hospital outpatient physical therapy department; approval received from an orthopaedic specialist to weightbear as tolerated or partial weightbear; skeletal maturity; reduced passive dorsiflexion range of motion in the fractured ankle (ie, at least 5° less than the contralateral ankle); willingness to be available for the 3 month followup period. Patients were excluded if it was not possible to perform the baseline assessment within 5 days of cast removal or if they had concomitant pathologies (eg, symptomatic osteoarthritis, stroke, or other lower limb fractures). All potentially eligible patients were invited to participate in the study. Approval to conduct the study was obtained from the Northern Sydney Health Human Research Ethics Committee, the Central Sydney Area Health Service Ethics Review Committee, and the University of Sydney Human Ethics Committee. All subjects gave written informed consent to participate.
Initial orthopaedic management (ie, surgical or nonsurgical) was at the discretion of the treating orthopaedic specialist and was not mandated by the study protocol. Subjects were randomly allocated to one of three physical therapy treatment programs: exercise only, exercise plus short-duration passive stretch, or exercise plus long-duration passive stretch. All groups were given a 4 week course of exercises. In addition, subjects in the short-duration stretch group completed 6 minutes of stretching per day, and subjects in the long-duration stretch group completed 30 minutes of stretching per day. More details about the intervention and methods can be obtained from the primary analysis of the data set.13 Because there were no between-group differences in the randomized trial, we combined the three groups for this investigation.
At baseline (ie, within 5 days of cast removal), 4 weeks, and 3 months the relationship between passive torque and ankle plantarflexion-dorsiflexion motion of the fractured ankle were quantified using an instrumented footplate.12 Testing involved the subject lying supine with the knee extended and the foot secured in the footplate. The subject was instructed to relax as much as possible and the footplate was rotated manually to produce sinusoidal stretches for 30 seconds. Maximum dorsiflexion was defined as the angle at which the heel just commenced to lift off the footplate, or the angle where the subject reported a moderate stretch or pain. Stretching velocity was always less than 10°/s. The target stretching amplitude was 40°, but this could not be achieved for some subjects at baseline.
The force signal (from a strain-gauge type load cell) and angle signal (from a rotary potentiometer) were sampled at 50 Hz. The passive torque applied to the foot was calculated by subtracting the weight torque of the footplate from the measured torque. The last 25 seconds of torque and angle data were smoothed using a running average with a 0.1 second time window. These torque versus displacement curves were averaged and an exponential function was fitted to the average curve.12 The equation was τ = ekθ+b, where passive torque (τ, in Nm) and ankle angle (θ, in degrees) were measured variables. The two other variables, a stiffness coefficient (k) that reflects the slope of the log of the passive torque versus displacement curve and a preload coefficient (b) that reflects the log of the passive torque required to achieve 0° of dorsiflexion, were estimated with regression. This equation fitted the torque versus displacement data well; the regression coefficient always exceeded 0.67, with a median of 0.97. Each subject's coefficients from curve fitting were used to calculate the torque corresponding to the peak dorsiflexion angle at baseline.
We have previously published data from an able-bodied sample for ankle flexibility12 using a method identical to the one used here. Data from immobilized ankles were compared with the able-bodied data. To compare the two groups for the torque corresponding to the peak dorsiflexion angle at baseline, each subject with ankle fracture was randomly paired with a subject from the able-bodied sample. At the ankle fracture subject's peak dorsiflexion value, the passive torque was calculated for each able-bodied subject using the subject's stiffness and preload coefficient values.
Only subjects completing all three assessments were included in the analysis. The three dependent variables analyzed were the stiffness coefficient, the preload coefficient, and the torque corresponding to the peak dorsiflexion angle at baseline. Dependent samples t tests were used to compare baseline to 4 week values and 4 week to 3 month values for each variable. The data obtained from subjects with ankle fracture at baseline, 4 weeks, and 3 months were compared with the normative data using two-sample t tests. The influence of initial orthopaedic management on recovery of flexibility was investigated by comparing fracture subjects with and without surgical fixation for each variable and at each time point using two-sample t tests. Because of the number of comparisons undertaken, statistics with a probability of less than 1% under the null hypothesis were considered significant.
We recruited 150 people with ankle fracture. Complete ankle flexibility data were available for 125 of these subjects (16 withdrew from the study and nine had incomplete data because of equipment failure or torque values below 2 Nm at baseline). Published data from 194 able-bodied subjects were also used.12 The majority of subjects with ankle fracture sustained their fractures in falls. Approximately half the group required surgical fixation and three-quarters had a Weber classification of B. Able-bodied subjects were younger and more likely to be male than the ankle fracture group (Table 1).
Recovery of ankle flexibility was incomplete 3 months after cast removal. Both the stiffness coefficient and the torque corresponding to the peak dorsiflexion angle at baseline decreased towards normal values during the 3 month recovery period (Fig 1). The mean stiffness coefficient at baseline was 0.088 Nm (SD, 0.023), decreasing to 0.074 Nm (SD, 0.020) at 4 weeks after cast removal (t = 8.50, p < 0.001), and reaching 0.065 Nm (SD, 0.018) at 3 months (t = 6.30, p < 0.001). Torque decreased from 2.29 Nm (SD, 0.21) at baseline to 2.14 Nm (SD, 0.21) at 4 weeks (t = 9.85, p < 0.001) and 2.05 Nm (SD, 0.21) at 3 months (t = 6.31, p < 0.001). While the subjects recovering from ankle fracture approached the stiffness (mean, 0.056 Nm; SD, 0.008) and torque (mean, 1.98 Nm; SD, 0.13) values of the able-bodied group, equivalence was never achieved (p < 0.01); mean values had not returned to normal by 3 months (p < 0.01).
In contrast, the log of the passive torque to achieve a neutral ankle position (ie, preload coefficient) did not vary with time post injury and was lower than the able-bodied group at all time points (Fig 1). The mean values for the ankle fracture group were not statistically different at baseline (mean, 1.49 Nm; SD, 0.47), 4 weeks (mean, 1.48 Nm; SD, 0.41), and 3 months (mean, 1.54 Nm; SD, 0.40). However, preload coefficients in the ankle fracture group were lower (p < 0.01) than in the able-bodied group (mean, 1.87 Nm; SD, 0.28) at all time points.
Subjects who had surgical fixation following ankle fracture had higher stiffness and preload coefficients and higher torque corresponding to the peak dorsiflexion angle at baseline than subjects with nonsurgical management (Table 2). Stiffness, preload and torque were substantially higher at baseline (p < 0.001, p = 0.005, p < 0.001, respectively) for subjects with surgical fixation. Stiffness values were also markedly higher at each follow up (p < 0.001 at 4 weeks, p = 0.002 at 3 months) and torque values were substantially higher (p < 0.001) at 4 weeks.
We documented recovery of passive ankle plantarflexion-dorsiflexion flexibility after cast immobilization for ankle fracture. Soon after cast removal passive stiffness and the torque corresponding to the peak dorsiflexion angle at baseline are substantially higher than an able-bodied group, particularly for people with surgical fixation. Stiffness and torque values progressively decrease towards able-bodied values over a 3-month period. Nonetheless, abnormal levels of stiffness and torque are still present 3 months after cast removal.
Increased passive stiffness after cast immobilization for ankle fracture has been previously documented in one small case series4 and a report of three cases.20 Passive stiffness at 0° of dorsiflexion 2 days after cast removal was higher in the fractured limb compared to the contralateral uninjured limb in people with severe fractures (n = 11), but not in people with less severe fractures (n = 19).4 High resting ankle stiffness 7 to 9 days after cast removal was noted in three cases, with stiffness decreasing over the next 20 days.20
The increase in ankle stiffness is reflected in a decreased range of joint motion. Passive range of motion is difficult to measure using a torque-controlled measurement procedure (ie, the gold standard)11 soon after cast immobilization for ankle fracture because some subjects tolerate only very low levels of passive torque due to pain. Our method of quantifying loss of range of motion allowed the peak passive torque to vary between subjects (range, 3-14 Nm).13 On average, there was 13° (SD, 9°) loss of dorsiflexion at baseline in our group,13 which is similar to the value reported by Chesworth and Vandervoort (mean 11° for more severe and 12° for less severe fractures).4
We quantified two passive torque variables: the preload coefficient (ie, the log of the passive torque required to achieve 0° of dorsiflexion) and the torque corresponding to the peak dorsiflexion angle at baseline. Decreases in these variables indicate after immobilization the passive torque-displacement curve of the ankle shifted to the right (ie, the ankle became more flexible). While the torque corresponding to the peak dorsiflexion angle is always higher than normal and decreases over the 3-month recovery period for our sample with ankle fracture (Fig 1), the same cannot be said for the preload coefficient. Preload was lower than normal at all time points (Fig 1), suggesting the flat portion of the passive torque-displacement curve becomes progressively longer during the recovery period.
The length of time required for humans to recover from immobilization-induced changes in passive ankle flexibility is not well documented.18,20 It has been suggested some muscle characteristics recover to normal values with 10 weeks of physical therapy after an 8-week period of immobilization.18 However, we found passive stiffness and torque were still higher than normal 3 months after cast removal in subjects who were immobilized after ankle fracture for an average of 6 weeks.
The critical stimulus for recovery of passive ankle plantarflexion-dorsiflexion flexibility after cast immobilization for ankle fracture is not clear.6 The primary analysis of this randomized trial revealed the addition of a program of passive stretches conferred no benefit over exercise alone for activity limitation or passive dorsiflexion range of motion.13 The weightbearing exercise prescribed to all groups, along with return to usual daily activities, may therefore have provided the primary stimulus for recovery.9,23,24
We note several limitations in our investigation. One limitation of our analysis is the use of normative data from able-bodied subjects who were younger and more likely to be male12 (Table 1). There is conflicting evidence about the effect of age on passive flexibility3,19 and males appear to have higher passive torque values than females.19 Three study design features may affect the generalizability of our data. First, there were no guidelines for the initial orthopaedic management (ie, surgical versus nonsurgical). Second, our sample may be biased towards more severe ankle fracture as we excluded patients who were not referred to physical therapy and who did not have plantarflexion con-tracture. Third, our followup period (ie, 3 months) was relatively short.
Our data suggest stiffness and the passive torque corresponding to the peak dorsiflexion angle at baseline were higher than normal, particularly for people with surgical fixation; this increase was apparent within 5 days of cast removal. Stiffness and torque progressively decrease over a 3 month period, but full recovery takes longer than 3 months.
The authors thank the physical therapists at Royal North Shore Hospital (particularly Deborah Taylor and Trish Evans) and Royal Prince Alfred Hospital (particularly Sandeep Gupta, Gavin Robertson and Julie Penn) who assisted in recruitment and provided the experimental treatments. We also thank Stephanie Lanzarone and Adrian Byak, who assessed outcomes.
1. Airaksinen O. Changes in posttraumatic ankle joint mobility, pain, and edema following intermittent pneumatic compression therapy. Arch Phys Med Rehabil
2. Akeson WH, Woo SL, Amiel D, Matthews JV. Biomechanical and biochemical changes in the periarticular connective tissue during contracture development in the immobilized rabbit knee. Connect Tissue Res
3. Chesworth BM, Vandervoort AA. Age and passive ankle stiffness in healthy women. Phys Ther
4. Chesworth BM, Vandervoort AA. Comparison of passive stiffness variables and range of motion in uninvolved and involved ankle joints of patients following ankle fractures. Phys Ther
5. Hancock MJ, Herbert RD, Stewart M. Prediction of outcome after ankle fracture. J Orthop Sports Phys Ther
6. Herbert RD. The prevention and treatment of stiff joints. In: Crosbie WJ, McConnell J, eds. Key Issues in Musculoskeletal Physiotherapy
. London: Butterworth-Heinemann; 1993:114-141.
7. Herbert RD, Balnave RJ. The effect of position of immobilisation on resting length, resting stiffness, and weight of the soleus muscle of the rabbit. J Orthop Res
8. Herbert RD, Crosbie J. Rest length and compliance of nonimmobilised and immobilised rabbit soleus muscle and tendon. Eur J Appl Physiol
9. Kauhanen S, Leivo I, Pettila M, Michelsson JE. Recovery of skeletal muscle after immobilization of rabbit hindlimb: a light microscopic study. APMIS
10. Michelson JD. Fractures about the ankle. J Bone Joint Surg Am
11. Moseley A, Adams R. Measurement of passive ankle dorsiflexion: procedure and reliability. Aust J Physiother
12. Moseley AM, Crosbie J, Adams R. Normative data for passive ankle plantarflexion-dorsiflexion flexibility. Clin Biomech (Bristol, Avon)
13. Moseley AM, Herbert RD, Nightingale EJ, Taylor DA, Evans TM, Robertson GJ, Gupta SK, Penn J. Passive stretching does not enhance outcomes in patients with plantarflexion contracture after cast immobilization for ankle fracture: a randomized controlled trial. Arch Phys Med Rehabil
14. Reynolds CA, Cummings GS, Andrew PD, Tillman LJ. The effect of nontraumatic immobilization on ankle dorsiflexion stiffness in rats. J Orthop Sports Phys Ther
15. Reznick AZ, Menashe O, Bar-Shai M, Coleman R, Carmeli E. Expression of matrix metalloproteinases, inhibitor, and acid phosphatase in muscles of immobilized hindlimbs of rats. Muscle Nerve
16. Tabary JC, Tabary C, Tardieu C, Tardieu G, Goldspink G. Physiological and structural changes in the cat's soleus muscle due to immobilization at different lengths by plaster casts. J Physiol
17. van Staa TP, Dennison EM, Leufkens HGM, Cooper C. Epidemiology of fractures in England and Wales. Bone
18. Vandenborne K, Elliott MA, Walter GA, Abdus S, Okereke E, Shaffer M, Tahernia D, Esterhai JL. Longitudinal study of skeletal muscle adaptations during immobilization and rehabilitation. Muscle Nerve
19. Vandervoort AA, Chesworth BM, Cunningham DA, Paterson DH, Rechnitzer PA, Koval JJ. Age and sex effects on mobility of the human ankle. J Gerontol
20. Weiss PL, Kearney RE, Morier R. Quantitative assessment of ankle joint dynamics during recovery from injury. Clin Biomech (Bristol, Avon)
21. Williams PE, Goldspink G. Connective tissue changes in immobilised muscle. J Anat
22. Woo SL, Buckwalter JA. Injury and Repair of the Musculoskeletal Soft Tissues
. 1988, Park Ridge, IL, American Academy of Orthopaedic Surgeons.
23. Zarzhevsky N, Carmeli E, Fuchs D, Coleman R, Stein H, Reznick AZ. Recovery of muscles of old rats after hindlimb immobilisation by external fixation is impaired compared with those of young rats. Exp Gerontol
24. Zarzhevsky N, Coleman R, Volpin G, Fuchs D, Stein H, Reznick AZ. Muscle recovery after immobilisation by external fixation. J Bone Joint Surg Br
25. Zarzhevsky N, Menashe O, Carmeli E, Stein H, Reznick AZ. Capacity for recovery and possible mechanisms in immobilization atrophy of young and old animals. Ann N Y Acad Sci