Improvements in balance after total hip replacement

M. Majewski, H. A. BischoffFerrari, C. Grneberg, W. Dick, J. H. J. Allum

From the University Hospital of Basel, Basel, Switzerland

M. Majewski, MD, Orthopaedic Surgeon Department of Orthopaedics Kantonsspital, CH-4410 Liestal, Switzerland. J. H. J. Allum, DSc, Professor Department of Otorhinolaryngology W. Dick, MD, Professor and Orthopaedic Surgeon Department of Orthopaedics University Hospital Basel, CH-4031 Basel, Switzerland. C. Grneberg, PhD, Research Leader Europa Fachhlochschule Fresenius University of Applied Sciences, Fachbereich Gesundheit Limburger Strae 2, 65510 Idstein, Germany. H. A. Bischoff-Ferrari, MD, MPH, Assistant Professor Department of Rheumatology and Institute of Physical Medicine and Rehabilitation University Hospital of Zurich, CH-8091 Zurich, Switzerland. Correspondence should be sent to Professor J. H. J. Allum; e-mail: jallum@uhbs.ch 2005 British Editorial Society of Bone and Joint Surgery doi:10.1302/0301-620X.87B10. 16605 $2.00 J Bone Joint Surg [Br] 2005;87-B:1337-43. Received 8 April 2005; Accepted 4 May 2005

We have investigated whether control of balance is improved during stance and gait and sitto-stand tasks after unilateral total hip replacement undertaken for osteoarthritis of the hip. We examined 25 patients with a mean age of 67 years (SD 6.2) before and at four and 12 months after surgery and compared the ndings with those of 50 healthy age-matched control subjects. For all tasks, balance was quantied using angular measurements of movement of the trunk. Before surgery, control of balance during gait and sit-to-stand tasks was abnormal in patients with severe osteoarthritis of the hip, while balance during stance was similar to that of the healthy control group. After total hip replacement, there was a progressive improvement at four and 12 months for most gait and sit-to-stand tasks and in the time needed to complete them. By 12 months, the values approached those of the control group. However, trunk pitch (forwards-backwards) and roll (side-to-side) velocities were less stable (greater than the control) when walking over barriers as was roll for the sit-to-stand task, indicative of a residual decit of balance. Our data suggest that patients with symptomatic osteoarthritis of the hip have marked decits of balance in gait tasks, which may explain the increased risk of falling which has been reported in some epidemiological studies. However, total hip replacement may help these patients to regain almost normal control of balance for some gait tasks, as we found in this study. Despite the improvement in most components of balance, however, the decit in the control of trunk velocity during gait suggests that a cautious follow-up is required after total hip replacement regarding the risk of a fall, especially in the elderly.

Most patients report considerable relief from pain and improved function after total hip replacement (THR)1-5 although the degree to which the procedure inuences post-operative balance has not been previously studied. Problems with balance and gait can occur in patients with severe osteoarthritis (OA) of the hip because of damage to proprioceptors, and they are further damaged as a result of capsular excision during THR. In addition, weakness of the abductors, leg-length inequality, a shortened lever arm, a restricted range of movement, and altered weight-bearing may impair balance.6,7 The functional outcome after THR may be assessed by measuring the time needed to complete specic gait and balance tasks.8-12 An alternative, albeit more time-consuming method, is to measure angular movements of the body with a movement analysis system.13,14 A simpler method, however, is to measure balance as trunk pitch (forwards-backwards) and roll (side-to-side) angles in addition to velocity dur-

ing stance, gait and sit-to-stand tasks. This last method is the one which we used.15,16 Our aim was to evaluate whether balance performance improved after THR for OA and, if so, whether it could return to normal.

Patients and Methods

Twenty-ve consecutive patients (11 women and 14 men) with a mean age of 67 years (SD 6.2) were examined prospectively. The main criterion for inclusion in the study was severe, concentric, unilateral OA of the hip (12 right, 13 left) and typical symptoms of degenerative joint disease. In all patients, the radiological degeneration was grade 4, according to Kellgren and Lawrence.17 Patients were tested on three occasions: pre-operatively, when their symptoms were most severe; and four months and one year after surgery. An adapted Harris hip score was used for pre- and post-operative clinical and radiological comparisons.18 Patients with bilateral OA of the hip were excluded, as were those with OA of the knee,
1337

VOL. 87-B, No. 10, OCTOBER 2005

1338

M. MAJEWSKI, H. A. BISCHOFF-FERRARI, C. GRNEBERG, W. DICK, J. H. J. ALLUM

40 Angle () a

Pitch 0 -10 0 40 Roll 1 2 3 Time (s) 4 5 6 7

Angle ()

Pitch 0 -10 0

Roll 1 2 3 4 5 Time (s)

40 Angle ()

Pitch c

0 -10 0

Roll 1 2 Time (s) 3 4

Fig. 1 Photograph showing the SwayStar balance system for measuring trunk pitch (forwards-backwards) and roll (side-to-side) angular movements during the stance and gait tasks. The belt was placed around the patient so that the sensors were positioned at level of L2/L3.

Fig. 2 Traces showing the improved ability of a typical patient with a THR to rise from a stool and walk three metres. Measurements were made during this get-up-and-go task a) before, and b) at four and c) at 12 months after surgery. Forward pitch and rightward roll are represented by an upward deflection of the traces. They indicate a significant reduction in the time needed to complete the task, as well as a greater degree of trunk flexion.

ankle or foot. No patient had undergone any surgery on the ipsilateral leg before this study, nor had any suffered from other health problems which could have affected gait and balance, such as spinal stenosis. All patients received the same type of THR, a Morscher press-t acetabular component (Zimmer, Winterthur, Switzerland) and a cemented MS-30 femoral component (Zimmer) inserted through a transgluteal approach with the patient in the supine position. All operations were performed either by or under the supervision of an experienced surgeon (WD) with regional anaesthesia being used in most cases. After surgery all patients were mobilised fully weight-bearing with crutches on the rst post-operative day. They could sit on a hip chair, but were not allowed to cross their legs for six weeks. A control group comprised 50 normal, healthy subjects, who were age- and gender-matched to each of the patients from a database of normal individuals.19 They had no complaints of vertigo or other problems affecting gait and balance. The subjects in both the study and control groups gave their written informed consent to participate in the investigation, which had received approval from the ethical review board of the hospital. Balance assessment device. Control of balance, in the form of trunk pitch (forwards-backwards) and roll (side-to-side) movements, was measured during the stance and gait tasks by the two angular velocity sensors (breoptic gyroscopes) of the SwayStar balance system (Balance International Innovations GmbH, Iseltwald, Switzerland), which was attached by a belt to the patient so that the sensors were at

the L2/3 level (Fig. 1). The sensors were connected by a tenmetre cable to a computer which sampled the angular velocity signals every 100 ms and numerically integrated the signals to yield angular displacement. The specied maximum baseline drift of the sensors was 36/hr or 0.01/s and the noise about this drift was 0.6/hr. Since the maximum task duration was 20 s, the maximum cumulative measurement error was 0.2. Stance and gait tasks. Patients were asked to perform a protocol of ve standard tasks derived from earlier publications.10,12 They consisted of one sit-to-stand, two gait and two stance tasks. For the sit-to-stand task, we performed the rst half of the timed get-up-and-go test.20,21 For this, the patient sat on a height-adjusted stool with their knees exed to 90. On command they stood up, without placing their hands on their knees in order to lever themselves forwards and they then walked a distance of three metres. For the rst gait task, the patients walked over a series of four low barriers, each 24 cm high and one metre apart. For the second, they walked up a set of stairs comprising two upward and two downward steps each 23 cm high, without handrails. For the rst stance task, the patients stood on both legs on a normal surface with their eyes open. For the second, they stood on both legs on a foam support surface with their eyes closed. The foam surface was 10 cm thick, 44 cm wide, 204 cm long and had a density of 25 kg/m2. For the stance tasks the patients stood without shoes for a maxiTHE JOURNAL OF BONE AND JOINT SURGERY

IMPROVEMENTS IN BALANCE AFTER TOTAL HIP REPLACEMENT

1339

Table I. Statistical differences (p values) for the 25 patient with a THR at the three time intervals and between the patients and the 50 normal subjects Variable (peak-to-peak amplitude) Duration Get-up-and-go task 4 months compared with pre-operatively 1 year compared with pre-operatively Pre-operative compared with normal 4 months compared with normal 1 year compared with normal Barriers task 4 months compared with pre-operatively 1 year compared with pre-operatively Pre-operative compared with normal 4 months compared with normal 1 year compared with normal Stairs task 4 months compared with pre-operatively Pre-operative compared with 1 year Pre-operative compared with normal 4 months compared with normal 1 year compared with normal Standing (eyes open) task Pre-operative compared with normal 4 months compared with normal 1 year compared with normal Standing (eyes closed) task Pre-operative compared with normal 4 months compared with normal 1 year compared with normal 0.001 (shorter)* 0.026 (shorter) 0.001 (smaller) 0.003 (smaller) 0.011 (smaller) Pitch angle () Pitch velocity 0.049 (faster) 0.001 (faster) 0.001 (slower) 0.011 (slower) Roll angle () Roll velocity(s)

0.013 (larger)

0.012 (larger)

0.002 (shorter) 0.001 (longer)

0.003 (larger) 0.004 (larger) 0.004 (larger) 0.002 (larger) 0.001 (larger) 0.004 (larger) 0.011 (larger) 0.023 (larger) 0.006 (larger)

0.033 (larger)

0.001 (shorter) 0.029 (longer) 0.025 (shorter) 0.045 (smaller) 0.022 (smaller) 0.005 (smaller) 0.001 (smaller) 0.009 (smaller) 0.006 (smaller) 0.001 (smaller)

0.042 (smaller)

0.001 (smaller) 0.001 (smaller) 0.026 (smaller) 0.023 (smaller) 0.041 (smaller)

0.001 (smaller)

0.034 (smaller) 0.009 (smaller)

* duration (shorter) implies, for example, if the four-month is compared with the pre-operative result, that the time to complete the task was shorter. Duration (longer) implies the converse. velocity (faster) implies, for the get-up-and-go task that trunk pitch velocity was faster at four months compared with pre-operatively. Velocity (faster) implies the converse. For other tasks, the term larger (or smaller) velocity is used. angle (smaller) implies, for example, that the amount of trunk movement was less than normal. Angle (larger) implies the converse.

mum of 20 s, if possible. Two physiotherapists stood beside the patient throughout the tests in order to prevent a fall. The duration of the test was the time to completion of the task or the time taken until a fall would have occurred but for the action of the physiotherapist. The task protocol was repeated three times for each patient. The rst occasion was between two and ve days before surgery. The second was performed four months later (mean 119 days, SD 9.2) and the third was at follow-up at 12 months (mean 391 days, SD 8.6). Validation of these tasks had already been performed.15 Analysis of data. The method of data analysis has been described previously.15,16,19 From each recording the maximum peak-to-peak excursions of pitch and roll angle and velocity were calculated to yield ve variables, with duration, to quantify each task. Before the statistical analysis, the peak-to-peak measurements were log transformed to change their distribution from Poisson-like to Gaussian.19 A Kolmogorov-Smirnov test was used to check this. A repeated measures analysis of variance was used to determine the effect of post-operative time on the measurement variables. Individual variables with a signicant time effect were investigated using Bonferroni tests, with signicance set at p 0.05, in order to determine if their mean values measured pre-operatively
VOL. 87-B, No. 10, OCTOBER 2005

and at four and 12 months, were different. In addition, the mean values for these time periods were also compared with those of the control group.

Results
Get-up-and-go task. The initial phase of the get-up-and-go

task is characterised by a rapid forward exion of the trunk which provides the momentum to stand and to start walking. Clearly, the inability to pitch the trunk forwards rapidly enough, as may occur with severe OA of the hip, compromises the ability to complete this task (Fig. 2). In the patients, pitch velocity increased after surgery at four (p = 0.049) and at 12 months (p = 0.001), approaching the value of the normal control group by 12 months. The difference between the patients and the control group was signicantly different pre-operatively (p = 0.001) and at four months (p = 0.011), but not at 12 months. Coincidental with this increase in pitch velocity, the signicant reduction in the amplitude of the pitch angle before surgery (p = 0.001) had also improved at four (p = 0.003) and 12 months (p = 0.011) post-operatively. However, forward trunk exion remained reduced after THR (p = 0.01, compared with controls) although pitch velocity had increased. The roll velocity, which was not signicantly abnormal before surgery (p > 0.05), increased slightly at four months (p = 0.05), and more so at 12 months (p = 0.012). This sug-

1340

M. MAJEWSKI, H. A. BISCHOFF-FERRARI, C. GRNEBERG, W. DICK, J. H. J. ALLUM

Trunk velocity (/s) 80 40 Pitch 0 -40 -80 -16

Roll Trunk angle ()

-8

16

12 8 Pitch 4 0 -4 -12 -8
Fig. 3a

-4

0
Fig. 3b

12
Fig. 3c Fig. 3d

Illustration of trunk pitch and roll velocity angle and for a typical patient a) before and b) at four and c) 12 months after THR while performing the barriers task, and d) for the normal control group. Each recording is shown as an x-y plot of either pitch versus roll angle (upper row) or pitch versus roll velocity (lower row). An envelope has been drawn around the excursions. The maximum peak-to-peak excursions in each direction were analysed. There is a reduction of pitch and roll angle and an increase in pitch and roll velocity with time.

Pre-operative a 180 160 140 Velocity (/s) 120 100 80 60 40 20 0 Roll b 30 25 20 Angle () 15 10 5 0 Roll

4 mths

12 mths

Normal control group

Pitch

Pitch Fig. 4

Bar charts showing the mean, and SEM values for a) the roll and pitch velocity and b) the roll and pitch angle for the barriers task. The arrow indicates significant (p < 0.05) changes in the mean with time. The mean value approaches the normal mean value, with velocities becoming greater than normal. The mean values are greater than median values because the measurements have an asymmetrical distribution across the sample populations1 (* statistically significant p < 0.05).

gested a higher, possibly unstable roll velocity at 12 months (p = 0.013) compared with the situation before operation. The time needed to complete the get-up-and-go-task improved signicantly after surgery. This was already obvious by four months (p = 0.001) and was maintained at 12 months (p = 0.026; Table I). Barriers task. The mean values for all the patients showed a progressive decrease in pitch and roll angles and a progressive increase in pitch and roll velocity at four and 12 months post-operatively for the barriers task (Fig. 3). These angles approached the values of the normal control group (Fig. 4). Pitch velocity increased signicantly at 12 months compared with pre-operatively (p = 0.003) reaching values exceeding the normal (p = 0.033). Similarly, roll velocity in the patients increased to a value beyond the normal at four months (p = 0.023) and remained so at 12 months (p = 0.006). Compared with the control group, measurements of the amplitude of the pitch and roll angle showed signicantly higher values before surgery and after four and 12 months (Table I). The time needed to complete the barriers task was significantly increased before surgery in the patients when compared with the control group (p = 0.001), but by the 12month follow-up was closer to that of the control group (p = 0.002). Thus, by 12 months the patients could complete the task as quickly as the normal group but their trunk
THE JOURNAL OF BONE AND JOINT SURGERY

IMPROVEMENTS IN BALANCE AFTER TOTAL HIP REPLACEMENT

1341

8 7 Duration (s) 6 5 4 3 2 1 0 Fig. 5 Bar charts showing the mean duration, and SEM values for the task of walking up and down stairs. The time needed to complete the task was considerably shorter at 12 months (arrow), when the patients were faster than the normal (*statistically significant p < 0.05). Normal control group Pre-operative 4 mths 12 mths

movements were larger and faster, indicative of residual instability (Table I). Stairs task. Walking up and down stairs involves clearing the foot over the height of the step, but also requires loadbearing capabilities to pull the body either up to the next step or to stabilise it when stepping down.

The mean duration of all participants for this task is illustrated in Figure 5. The duration reduced signicantly after one year compared with the normal control group (p = 0.025) and showed a signicant reduction in the time needed to complete the task compared with pre-operatively (p = 0.001). We also observed small reductions in the pitch angle of the trunk after four (p = 0.045) and 12 months (p = 0.022), possibly caused by the effects of improved movements of the hip on movements of the trunk, although these changes were not as signicant as those for the get-up-and-go tasks (Table I). Stance on two legs with the eyes open or closed. At all time points there was no decit of balance during the stance tasks. The pitch and roll angle and the pitch and roll velocity for normal standing with the eyes open were signicantly less than those of the control group at all time points. Less roll sway was seen when standing with eyes closed on a foam support surface (Fig. 6). Thus, roll angle and roll velocity appeared to be lower in the patients than in the control group under both stance conditions, suggesting that the patients stood more stify than the control group (Table I).

12 10 Velocity (/s) 8 6 4 2 Pre-operative 0 Roll Pitch 4 mths 12 mths

Normal control group

3 Angle ()

0 Roll
Fig. 6 Bar charts showing the mean values for trunk sway a) velocity and b) angle measurements and SEM values for the stance task when standing on a foam support surface with the eyes closed (*statistically significant p < 0.05).
VOL. 87-B, No. 10, OCTOBER 2005

Pitch

1342

M. MAJEWSKI, H. A. BISCHOFF-FERRARI, C. GRNEBERG, W. DICK, J. H. J. ALLUM

Clinical and radiological ndings. Pre-operatively, all patients

had concentric arthritis of the hip with greatly reduced joint space and sclerosis of subchondral bone. The mean Harris hip score was 42 (23 to 52). At follow-up at one year this had risen to 95 (71 to 100). There were no local or systemic complications and no evidence of loosening of the components. Comparison of the pre- and post-operative radiographs showed no major differences in parameters such as the position of the centre of the hip or the abductor moment. Pre-operatively, most patients had a slight limblength discrepancy, with a mean of 0.5 cm (0 to 1) and all complained of pain during daily living. Post-operatively two patients had a limb-length discrepancy of 0.5 cm and two of 1 cm. Six complained of slight, occasional pain which did not compromise daily living.

Discussion
Our study has shown that balance during everyday gait and sit-to-stand tasks is signicantly altered by severe OA of the hip. Patients undergoing THR were assessed before surgery, and at four and 12 months, in order to track improvements in the control of balance after surgery. Their performance was also compared at each time point with that of the agematched healthy control group by using measurements of trunk angular movement. In individuals undergoing THR, balance while walking over low barriers and while getting up from a stool improved signicantly after surgery and approached the values of the age-matched control group at four and 12 months. For these tasks abduction, exion of the hip and strength are critical. Almost normal functional performance at 12 months suggested that surgery successfully restored joint mobility and control in these patients. However, despite the improvement in most balance measurements for gait and sit-to-stand tasks, roll velocities exceeded values of the normal control group at follow-up at 12 months. Similarly, the roll angular amplitudes remained larger than those of the control group for the barriers task. This suggests that balance for these tasks may either need more time to recover or that there may be some impairment which cannot be restored after THR. In our earlier studies on other groups, abnormal roll during gait tasks correlated with the incidence of falls.15,22 Initially, balance while standing appeared not to be compromised in patients with severe OA of the hip. Roll during both stance tasks was less than for the control group and did not change after surgery. In both stance tasks, roll angles and roll velocities when standing were lower than those in the control group suggesting that movements of the trunk and pelvis were restricted, possibly by general stiffness. One interpretation of these results is that it is preferable to assess balance by means of gait rather than by standing tasks in the follow-up of patients after THR, since the gait tasks were the most responsive to THR. In contrast to the remaining decits of balance in the patients, the time needed to perform different tasks gener-

ally improved towards that of the healthy control group by the follow-up at 12 months. Therefore, timing assessments during functional tasks may be of limited value if other aspects of the control of the balance are missed. This may be especially important since a large number of the patients undergoing THR are over the age of 65 years, an age when falls begin to occur frequently.7 It is currently unknown if the risk of falling in individuals with OA of the hip is reduced after THR. Balance has not been previously tested during functional tasks of the lower limbs in patients undergoing THR. Our ndings of improved function are supported, however, by several studies which addressed alterations in the electromyographic (EMG) patterns and changes in muscle strength in the leg muscles in patients with THR. Long et al,23 for example, examined such patients before and after operation by both EMG and gait analysis. Pre-operatively, they found that the EMG patterns during gait were abnormal, caused by an absence of activity in gluteus medius, tensor fascia lata and rectus femoris during walking. Two years after operation these abnormal patterns had disappeared. Murray et al24 also found that there was a signicant improvement in almost all aspects of functional gait performance two years after surgery. Similarly, Shih et al25 showed improved muscle strength in patients undergoing THR at six months and at one year after operation. These ndings support our assumption that post-operative increases in muscle strength may well underlie the functional improvements in task performance which we saw in the patients with THR. The improvements in balance which we observed could, however, be the result of increased strength and/or joint exibility. The increased speed with which patients were able to perform gait and sit-to-stand tasks at four and 12 months after surgery was probably because of greater muscle strength.26 Increased joint stiffness is a common clinical sign in patients with OA of the hip and is related to balance.7,13,24 Our analysis showed that patients with severe OA of the hip have less roll than normal subjects during stance. This was unaltered after THR and suggested that these patients were standing more stify than the control group even after surgery. We have previously addressed the question of stiffness of the hip in an experiment in young normal subjects who wore a cast to give articial stiffness at the hip and pelvis.27 We found the young subjects had faster roll velocities, but unchanged trunk pitch and roll angles, when correcting disturbances in stance. These ndings were similar to the balance decit of our patients when performing the barriers and get-up-and-go tasks. Consequently, our results in the articially-stiffened young subjects may explain some of the inability of patients with a THR to control their movements adequately because of a generally increased stiffness at the hip and pelvis.27,28 We, therefore, suggest that increased stiffness in patients with severe OA of the hip may lead to better control of balTHE JOURNAL OF BONE AND JOINT SURGERY

IMPROVEMENTS IN BALANCE AFTER TOTAL HIP REPLACEMENT

1343

ance in the form of less pitch and roll, while standing on two legs. By contrast, a decit of balance occurs if the trunk cannot hinge into an upright position with respect to the pelvis and legs while walking. The trunk cannot counterrotate as much because of the underlying stiffness, and must sway more to follow the course of leg abduction. In summary, the larger roll velocities seen during gait tasks suggest that overall stiffness is only partly reduced after surgery, while increased strength, reduced pain and greater exibility of the hip after THR lead to improved control of the trunk and balance. Future studies need to address how the larger roll velocities relate to both the risk of a fall and long-term function in patients undergoing THR. A patients ability to balance is related to his or her jointposition sense, derived primarily from receptors within muscle-tendon units and the joint capsule receptors. It is possible that this sense is compromised in patients with articial joints.29 Indeed, our ndings suggest that patients undergoing THR approach the levels of the control group after surgery in most tasks except for their control of trunk roll. It is unlikely that the improvements which we noted in functional performance were simply because of a learning effect. The periods of four and 12 months as inter-test intervals tend to preclude such an effect. Normal subjects, when retested within a shorter interval of three weeks, show no signicant learning effect.16 Our study acts as a basis for future investigations into the benets of different surgical techniques and implants but, more importantly, for the ongoing assessment of functional improvement after THR and any potential risk of a fall. The use of simple body-worn angular velocity transducers to quantify control of the trunk, and thereby performance of gait and balance, has proved to be very useful for this purpose.
This work was supported by grants from the Swiss National Science Foundation (31-59319.99) and the Free Academic Society of Basel to Professor J. H. J. Allum. We would like to thank H. U. Schlpfer, R. Akos and the team of physiotherapists at the Felix Platter Hospital in Basel for their co-operation in carrying out the protocols for this study. No benets in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

References
1. Felson DT, Zhang Y. An update on the epidemiology of knee and hip osteoarthritis with a view to prevention. Arthritis Rheum 1998;41:1343-55. 2. Guccione AA, Felson DT, Anderson JJ, et al. The effects of specic medical conditions on functional limitations of elders in the Framingham study. Am J Public Health 1994;84:351-8. 3. Liang MH, Bischoff HA, Roos EM. Outcome assessment in osteoarthritis: a guide for research and clinical practice. In: Brandt KD, Doherty M, Lohmander LS, eds. Osteoarthritis. Second ed. New York: Oxford University Press, 2003:381-90.

4. Harris WH, Sledge CB. Total hip and total knee replacement (2). N Engl J Med 1990;323:801-7. 5. Mahomed NN, Barrett JA, Katz JN, et al. Rates and outcomes of primary and revision total hip replacement in the United States medicare population. J Bone Joint Surg [Am] 2003;85-A:27-32. 6. Nallegowda M, Singh U, Bhan S, et al. Balance and gait in total hip replacement: a pilot study. Am J Phys Med Rehabil 2003;82:669-77. 7. Skinner HB. Pathokinesiology and total joint arthroplasty. Clin Orthop 1993;288: 78-86. 8. El-Kashlan HK, Shepard NT, Asher AM, Smith-Wheelock M, Telian SA. Evaluation of clinical measures of equilibrium. Laryngoscope 1998;108:311-19. 9. Gill-Body KM, Krebs DE, Parker SW, Riley PO. Physical therapy management of peripheral vestibular function: two clinical case reports. Phys Ther 1994;72:129-42. 10. Shumway-Cook A, Horak FB. Assessing the inuence of sensory interaction on balance: suggestion from the eld. Phys Ther 1986;66:1548-50. 11. Soderman P, Malchau H. Is the Harris hip score system useful to study the outcome of total hip replacement? Clin Orthop 2001;384:189-97. 12. Tinetti ME. Performance-orientated assessment of mobility problems in elderly patients. J Am Geriatr Soc 1986;34:119-26. 13. Arnold W, Schliebe G. Gait analysis in patients with total hip endoprosthesis: part II: pre- and postoperative gait analysis. Z Gesamte Inn Med 1992;47:15-20 (in German). 14. Ferrigno G, Pedotti A. Elite: a digital dedicated hardware system for movement analysis via real-time TV signal processing. IEEE Trans Biomed Eng 1985;32:943-50. 15. Adkin AL, Allum JHJ, Bloem BR. Trunk sway measurements during stance and gait tasks in Parkinsons disease. Gait and Posture 2005: in press. 16. Allum JH, Adkin AL. Improvements in trunk sway observed for stance and gait tasks during recovery from an acute unilateral peripheral vestibular decit. Audiol Neurootol 2003;8:286-302. 17. Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis 1957;16:494-502. 18. Harris WH. Traumatic arthritis of the hip after dislocation and acetabular fracture: treatment by mold arthroplasty: an end-result study using a new method of result evaluation. J Bone Joint Surg [Am] 1969;51-A:737-55. 19. Gill J, Allum JH, Carpenter MG, et al. Trunk sway measures of postural stability during clinical balance tests: effects of age. J Gerontol A Biol Sci Med Sci 2001;56: 438-47. 20. Bischoff HA, Stahelin HB, Monsch AU, et al. Identifying a cut-off point for normal mobility: a comparison of the timed up and go test in community-dwelling and institutionalised elderly women. Age Ageing 2003;32:315-20. 21. Podsiadlo D, Richardson S. The timed up & go: a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 1991;39:142-8. 22. De Hoon EWJ, Carpenter MG, Salis C, et al. Quantitative assessment of the stops walking while talking test in the elderly. Arch Phys Med Rehab 2003;89: 838-42. 23. Long WT, Dorr LD, Healy B, Perry J. Functional recovery of noncemented total hip arthroplasty. Clin Orthop 1993;288:73-7. 24. Murray MP, Gore DR, Brewer BJ, Mollinger LA, Sepic SB. Joint function after hip arthroplasty: a four-year follow-up of 72 cases with Charnley and Muller replacements. Clin Orthop 1981;157:119-24. 25. Shih CH, Du YK, Lin YH, Wu CC. Muscular recovery around the hip joint after total hip arthroplasty. Clin Orthop 1994;302:115-20. 26. Luepongsak N, Amin S, Krebs DE, McGibbon CA, Felson D. The contribution of type of daily activity to loading across the hip and knee joints in the elderly. Osteoarthritis Cartilage 2002;10:353-9. 27. Grneberg C, Bloem BR, Honegger F, Allum JHJ. The inuence of articially increased hip and trunk stiffness on balance control in the pitch and roll planes. Exp Brain Res 2004;157:472-85. 28. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med 1988;319:1701-7. 29. Skinner HB, Barrack RL, Cook SD, Haddad RJ Jr. Joint position sense in total knee arthroplasty. J Orthop Res 1984;1:276-83.

VOL. 87-B, No. 10, OCTOBER 2005