Share this article on:

Individuals With Total Knee Arthroplasty Demonstrate Altered Anticipatory Postural Adjustments Compared With Healthy Control Subjects

Venema, Dawn M. PT, PhD; Karst, Gregory M. PT, PhD

Journal of Geriatric Physical Therapy: April/June 2012 - Volume 35 - Issue 2 - p 62–71
doi: 10.1519/JPT.0b013e3182353ee4
Research Reports

Background: Persons with total knee arthroplasty (TKA) have many impairments that may compromise postural control. Most work examining postural control following TKA has focused on static and reactive postural control.

Purpose: The purpose of this study was to (1) compare anticipatory postural adjustments (APAs) between individuals with TKA and healthy controls; and (2) identify possible pre- to postoperative changes in APAs in those undergoing TKA.

Methods: Ten individuals planning TKA and 10 healthy age- and sex-matched controls were recruited. During a standing reaching task, lower extremity muscle activity was measured using electromyography (EMG) onsets and normalized EMG amplitudes, and center of pressure (COP) excursion was measured via a force platform. Other outcome measures included isometric strength of the knee flexors and extensors. Individuals in the TKA group were tested preoperatively, and at 3 and 6 months postoperatively. Controls were also measured 3 times over 6 months.

Results: There were no pre- to postoperative differences in lower extremity EMG onsets, normalized EMG amplitudes, or COP excursion in those with TKA. When compared to controls, individuals with TKA demonstrated lower EMG amplitudes of the vastus lateralis and biceps femoris, whereas EMG onsets and COP excursion did not differ. Individuals with TKA demonstrated lower knee extension torque.

Discussion and Conclusions: It seems that the surgery itself did not alter APAs among individuals with TKA. Potential contributors to the differences in EMG amplitudes in those with TKA compared to controls, such as impaired neural activation or efforts to reduce stress on the involved knee joint, need further investigation.

1Division of Physical Therapy Education, University of Nebraska Medical Center, Omaha.

2School of Allied Health Professions, University of Nebraska Medical Center, Omaha.

Address correspondence to: Dawn M. Venema, PT, PhD, Division of Physical Therapy Education, University of Nebraska Medical Center, Omaha, NE 68198-4420 (

This study was financially supported by the George Blanton Research Grant Program of the Nebraska Foundation for Physical Therapy.

Previous dissemination of this study: Portions of these data were presented as a scientific poster at Physical Therapy 2008: The Annual Conference and Exposition of the American Physical Therapy Association, San Antonio, TX, June 11–14, 2008.

The authors disclose no conflicts of interest.

Back to Top | Article Outline


Knee osteoarthritis (OA) has been found to compromise balance.15 Although the mechanism of balance impairment in persons with knee OA is not fully understood, impairments in strength and proprioception may play a role.35 In addition, some have suggested that decreased activity levels,5 pain,3 and joint instability4 could impact postural control in persons with knee OA.

Many characteristics of knee OA that may negatively impact balance persist in individuals after total knee arthroplasty (TKA), a common intervention for knee OA, particularly for individuals in their seventh and eighth decades of life.6 Deficits in lower-extremity strength and muscle activation are often present after TKA79 and could affect the body's ability to control the center of mass and use appropriate motor strategies for balance.10,11 Proprioceptive deficits in the affected knee joint have been observed in individuals with TKA12,13 and could impact sensory postural control strategies. Limitations in physical function following surgery1416 may prevent individuals from practicing postural control skills during daily tasks. Furthermore, falling can be an issue for individuals with TKA. One prospective observational study reported a 24.2% postoperative fall rate for individuals with TKA during the first year after surgery, with the rate as high as 45.8% for individuals identified as fallers prior to surgery.17 Thus, individuals with TKA can experience not only various impairments and activity restrictions that may impact balance, but also falls, making balance assessment in these individuals worthy of consideration.

Given the impact that TKA could have on balance, some investigators have explored whether or not deficits in balance or postural control strategies exist after TKA. Swanik et al18 reported an improved ability to balance on an unstable platform several months after surgery relative to presurgical performance, but did not include a control group to determine how performance compared with those without TKA. Gauchard et al19 found that although static sway improved over time after surgery, deficits persisted in patients with TKA relative to a control group, particularly in situations of altered somatosensory input. Alternatively, McChesney and Woollacott20 found no difference in static sway between subjects with TKA and a control group. McChesney and Woollacott20 and Gage et al21,22 studied how subjects with TKA and control subjects reacted to sudden movement of a platform upon which subjects were standing. McChesney and Woolacott20 noted no group differences in lower extremity electromyography (EMG) onset latencies with this task. Gage et al did note group differences in EMG onset latencies for platform movements in the frontal plane22 but not in the sagittal plane.21 Gage et al21,22 also reported lower EMG amplitudes in individuals with TKA. Thus, contingent on the type of task, dependent variables, and study design, postural control may or may not be altered in persons with TKA.

The studies mentioned previously involved assessment under quiet stance conditions (static postural control) or in response to disturbances imposed on the individuals (reactive postural control). None assessed individuals in response to self-initiated movements requiring postural adjustments prior to the movement (anticipatory postural control). Daily life is filled with voluntary movements that potentially destabilize the body, such as reaching outside one's base of support, making consideration of balance during these types of tasks worthwhile. Voluntary movements destabilize the body due to changes in limb and body alignment that shift the center of gravity, and through dynamic forces induced in linked body segments.23 Anticipatory postural adjustments (APAs) are thought to reduce displacement of the center of gravity when the trunk or upper extremities move over a stable base of support.23,24 APAs can be measured with EMG to assess the patterns of muscle activity which precede initiation of voluntary movement, and with force platforms to examine movement of the center of pressure (COP). Other researchers have noted deficits in APAs when studying standing reaching movements in individuals with multiple sclerosis,25,26 or with standing arm flexion in individuals with Parkinson disease,27 and cerebrovascular accident.28,29 It is possible that other chronic conditions such as OA could also lead to impaired APAs.

Therefore, the purpose of this study was to assess APAs during a standing reaching task in adults pre- and post-TKA and to compare them to a healthy age- and sex-matched control group. We hypothesized that after surgery, APAs would become more similar to those of an age- and sex-matched control group. We also hypothesized that despite any improvements, APAs would continue to be altered relative to a control group.

Back to Top | Article Outline


Study Design

This study was a 2-group, repeated measures design. Participants in the TKA group completed 3 visits over several months. A preoperative visit was held within 1 month before surgery. Postoperative sessions were held 3 and 6 months after surgery. Participants in the control group were also tested 3 times over 6 months at the same time intervals as individuals in the TKA group.

Back to Top | Article Outline


Participants in the TKA group were recruited through local orthopedic physician offices and hospitals, whereas those in the control group were recruited from the community. The study protocol was approved by the University of Nebraska Medical Center institutional review board. All subjects provided written informed consent to participate in the study.

Inclusion criteria for participants in the TKA group were (1) a diagnosis of knee OA and (2) a planned unilateral bi- or tri-compartmental TKA. No restrictions were made regarding surgeon, prosthesis type, or surgical approach. Inclusion criteria for control subjects were (1) no medical treatment for lower extremity OA (defined as a history of joint injections or current prescription medications) and (2) no lower extremity pain during the experimental tasks. Control participants were matched to members of the TKA group on the variables of age (±2 years) and sex. Exclusion criteria for both groups were (1) neurological disease, (2) diabetes mellitus, (3) rheumatoid arthritis, (4) osteoporosis, (5) history of lower extremity or spine injury currently limiting their mobility, (6) inability to do the experimental task due to limitations in upper extremity motion or the need for an assistive device, and (7) possible vestibular dysfunction, as screened with the Sensory Organization Test (SOT).30

Back to Top | Article Outline


Participants wore the same pair of walking shoes for all tasks during each testing session. Foot position was kept constant for all experimental tasks requiring stance. A single tester collected all data at all sessions to enhance reliability of procedures. Participants performed experimental tasks 1 to 4 listed later during all test sessions. Please note that additional detail about test procedures and analysis of biomechanical data is provided in the appendix.

Back to Top | Article Outline

1. Sensory Organization Test

The SOT was conducted using the Equitest System (NeuroCom® International, Inc, Clackamus, Oregon). If a participant's average scores on conditions 5 (eyes closed, sway-referenced support) or 6 (eyes open, sway-referenced support and surround) of the SOT fell below the 5th percentile of an age-matched sample,31 they were excluded from the remainder of the study. Intraclass correlation coefficients for test-retest reliability of conditions 5 and 6 of the SOT are reported as 0.68 and 0.64, respectively.32 Sensitivity of the SOT for detecting vestibular disorders has been reported as 40%, while specificity was found to be greater than 90%.33

Back to Top | Article Outline

2. Reaching Task

A standing reaching task was used to determine the participants' ability to perform a voluntary movement requiring an APA. This particular task has been previously used to study APAs in healthy adults of various ages,34,35 and in adults with multiple sclerosis.25,26 Individuals reached to a target placed at shoulder height and at distances of 40% and 80% of their maximal reach distance, measured while participants kept their feet flat on the ground. These distances were chosen as ones that required the individuals to reach outside their base of support, but provided a contrast between an easier (40%) and more difficult (80%) reach distance. Participants completed 5 reaches to each distance.

A triaxial piezoelectric accelerometer (output ± 5g, 400 mV/g; model TSD 109C, BIOPAC Systems, Inc, Santa Barbara, California) attached to the patient's wrist was used to determine onset of arm movement during the reaching task. The accelerometer also provided data on kinematic variables including peak wrist acceleration, time to peak wrist acceleration, and total movement time. These variables were collected because differences in movement kinematics could have an impact on EMG and COP variables.24,28,29

Electromyography data were collected from the surgical limb of participants in the TKA group and the corresponding limbs of the control participants. Electromyography data were recorded from the vastus lateralis (VL), biceps femoris (BF), tibialis anterior (TA), and lateral gastrocnemius (LG) using bipolar surface EMG electrodes with onsite preamplifiers (gain 290; Model: MA-110 by Motion Control, Inc, Salt Lake City, Utah) to determine the onsets and amplitudes of activity from these muscles during the reaching task. Electrode placement was kept constant across sessions by using standard methodology similar to that proposed by Zipp.36

During the reaching task, participants stood on a strain gauge force platform (Advanced Mechanical Technology, Inc, Watertown, Massachusetts) to collect the vertical ground reaction force (Fz) and the moment of force about the mediolateral axis (Mx). Center of pressure in the sagittal plane (COPy) was calculated using the following equation: COPy = Mx/Fz.

Back to Top | Article Outline

3. Gait

To normalize EMG amplitudes from the reaching task, reference EMG data were obtained as participants walked at a self-selected pace, a method of EMG normalization found to be reliable across sessions in young, healthy adults.37 EMG data were collected from the same muscles as the reaching task.

Back to Top | Article Outline

4. Isometric Strength Testing

A strain gauge force transducer (227 kg capacity; Interface Inc, Scottsdale, Arizona) was used to measure strength of quadriceps and hamstrings with participants in a seated position. The surgical lower extremity was tested for individuals in the TKA group, and the matching limb was tested for individuals in the control group. Participants performed 3 repetitions for each muscle group, with a 1-minute rest between contractions.

Back to Top | Article Outline

Surgery and Rehabilitation Interview: Participants in TKA Group Only

During a follow-up phone call 1-month postsurgery and at the 3- and 6-month test sessions, data about the participants' surgery and postsurgical rehabilitation were collected via interview.

Back to Top | Article Outline

Statistical Analysis

For EMG amplitudes and onsets, as well as COPy and kinematic variables, a mixed effects analysis of variance model was used to fit a repeated measures effect on the individual subject, which was nested within a random pair effect. Time and group were modeled as fixed main effects, and the interaction of time and group was also considered. Tukey's method was used for post hoc pair-wise analysis if a significant main effect of time was found.

To compare groups on descriptive data (age, weight, height, foot length), paired t tests were used. Comparisons between groups and over time for isometric strength data were done using a 2-way repeated measures analysis of variance. Only 9 pairs of participants were analyzed for isometric strength as this test was added after the first pair of participants had already begun the study. The level of significance for all statistical tests was set to α ≤ .05.

Back to Top | Article Outline


Descriptive Data

The Table provides descriptive data on each group. With the exception of sex and surgical limb, values are presented as means (standard deviations). Groups did not differ on the variables of sex, age, height, weight, or foot length.

Back to Top | Article Outline

Reaching Task

Figure 1 illustrates one participant's data from the reaching task. The data represent the average of 5 trials to the 40% reach distance from a participant in the TKA group. This figure shows the anticipatory nature of the TA and VL onsets with this postural control task.

Figure 2 compares groups on the variable of EMG onset timing for reaches to both distances. There were no changes in EMG onsets for any muscle over time in either group, so data presented are collapsed over time. There were no interactions between group and time. There were no between group differences in onsets for any muscles with the exception of the LG for the 40% reach distance (P = .0015). In this case, the LG was activated earlier for individuals with TKA.

Figure 3 compares groups on the variable of normalized EMG amplitudes for reaches to both distances. There were no changes in normalized EMG amplitudes for any muscle over time in either group, so data presented are collapsed over time. There were no interactions between group and time. There were no between group differences in normalized EMG amplitude for the TA or LG to either reach distance, but subjects with TKA had lower normalized EMG amplitudes for the VL and BF for both reach distances (all P values < .02).

For COPy excursion and kinematic data, there were no group or time effects, nor group ÷ time interactions for any of these variables for either reach distance.

Back to Top | Article Outline

Isometric Strength Testing

Figure 4 depicts a between group difference (P = .025), with subjects having TKA demonstrating less knee extension torque, but no significant change over time, nor a group ÷ time interaction. Knee flexion torque did not significantly differ between groups, over time, nor was there a group ÷ time interaction.

Back to Top | Article Outline

Surgery and Rehabilitation Interview: Participants in TKA Group Only

Participants from 2 hospitals and 4 surgeons took part in this study. Mean length of hospital stay was 3.9 days. While hospitalized, all participants received physical therapy (PT) for transfer and gait training. All but 2 were instructed in standard knee range of motion and isometric quadriceps exercises. After hospital discharge, 8 participants had home health PT, and 9 attended outpatient PT. The number of outpatient visits varied widely, ranging from 2 to 36. Outpatient PT interventions generally included progressive resistive exercise, electrical stimulation for muscle reeducation, stretching and range of motion exercises, gait progression, and stationary bicycling. Two participants participated in aquatic therapy. With the exception of training upright mobility and strengthening exercises in the standing position, no subjects reported performing any exercises designed specifically to address balance.

Back to Top | Article Outline


Anticipatory Postural Adjustments in Subjects With TKA

Our hypothesis that APAs in participants post-TKA would be more similar to controls was not supported as there were no changes over time in any of our dependent variables. However, our hypothesis that APAs in individuals with TKA would be altered relative to a control group was partially supported, as some group main effects were found. Most notably, participants with TKA had lower EMG amplitudes for the VL and BF than the control group, a finding that held across both reach distances. These differences cannot be explained by group differences in the kinematic performance of the movement.

Back to Top | Article Outline

EMG Amplitude

One explanation for the decreased EMG amplitude of the VL and BF in the TKA group is that these participants were adopting a strategy to minimize stress on their knee joint. Individuals with TKA have been found to perform proportionally less work with the involved lower extremity during a step-up task,38 and demonstrate less quadriceps EMG activity in the involved limb during sit to stand,39 when comparing to the uninvolved limb. Gage et al,21,22 who assessed reactive postural control in persons with TKA, found less EMG amplitude in individuals with TKA compared to controls, and suggested that those findings were consistent with a strategy to reduce stress on the postsurgical knee.

Decreased EMG amplitude in the TKA group may also be related to the finding that participants with TKA generated less torque at the knee than the control group, although the difference in knee flexion torque was not statistically significant. Muscle force is related in part to the activation of motor units to that muscle: the number of active motor units recruited, and the rate at which the motor neurons discharge action potentials.40 Therefore, decreased torque with isometric testing could be partially due to reduced neural activation of the muscle. Although it is impossible to know if reduced neural activation led to our participants' reduction in torque without performing a twitch superimposition technique,41 other investigators7,8 have previously reported reductions in neural activation of the quadriceps femoris when measuring knee extension torque in individuals with TKA. Reduced neural activation might also manifest itself as decreased EMG amplitude during postural control tasks, as seen in our study.

Back to Top | Article Outline

EMG Onset Latencies

Other investigators have noted delayed latencies in lower extremity musculature during APAs when comparing control subjects with patient populations, such as persons with multiple sclerosis,26 and cerebrovascular accident.28,29 The majority of latencies in our subjects were not different between groups, particularly in those muscles serving an anticipatory role, the TA and VL. This discrepancy could be explained by the fact that those studies that found delayed onsets included individuals with central nervous system involvement. Central nervous system involvement would impact the feed forward postural control strategies that lead to activation of the TA and VL in advance of the arm movement.

Back to Top | Article Outline

COPy Excursion

We found no group differences in COPy excursion for either reach distance. Using the same reaching task, Karst et al25 found that high-functioning individuals with multiple sclerosis demonstrated less COPy excursion than control subjects, likely as a compensatory strategy for preclinical balance deficits. Our finding of no group differences in COPy excursion suggests that group differences for EMG amplitudes were not great enough to have a significant effect on COPy excursion. One might expect that decreased activation of postural muscles would lead to less control of the COPy, but this was not the case.

Back to Top | Article Outline

Limitations and Strengths of the Study

In studies of balance and postural control, perhaps the most clinically meaningful outcome is whether or not an individual falls in daily life. Although we did find differences in APAs between groups, it is not clear if these differences have any functional relevance. In the 6-month follow-up period of our study, no participants in either group reported a fall, though falls were not formally tracked. The use of a longer follow-up period and formal tracking of falls would have been useful additions to this study to determine clinical significance. Also, future studies might consider comparing APAs in a cross-sectional analysis of persons with TKA both with and without a history of falls.

Range of motion was not measured in this study, but could have had an impact on the EMG data collected, particularly if knee angle differed between groups or within groups over time. The knee joint was near full extension during the reaching task, so it is possible that if individuals in the TKA group were lacking extension either before or after surgery, we could have been measuring EMG with their knees at slightly different angles than the control group, or at slightly different angles over time. Even within the control group, it is possible that there were slight variations in knee angle between subjects due to varying strategies between individuals, rather than differences in available range of motion.

An a priori sample size calculation was not done for this study, as we had no pilot data to provide us with effect sizes or measures of variance for our variables. We did find a statistically significant difference in EMG amplitudes for the VL and BF and for knee extension strength, so statistical power does not appear to be problematic for these variables. In addition, the group differences in EMG amplitudes held over both reach distances, giving weight to these findings. However, our study may have been underpowered for some of our other variables.

Experimental studies must strike a balance between internal and external validity. While limiting generalizability of the results, our stringent enrollment criteria likely led to better internal validity. By eliminating participants who may have had balance dysfunction because of other causes such as diabetes mellitus, neurological disease, or vestibular deficits, we can be more confident that the findings in this study are due to knee OA and TKA, not another disease process. On the contrary, the field from which participants were recruited was broad, leading to less experimental control. Subjects from 4 surgeons and 2 hospitals were enrolled. Likewise, there was no standardization of the rehabilitation in which the subjects took part. However, this variation in treatment makes our results more generalizable.

Study strengths also exist in data collection and study design. The same individual was present for all data collection, and completed all data analyses, likely leading to improved reliability of these aspects of the study. This study also combined a comparison with a control group matched on age and sex, with a longitudinal follow-up to compare performance pre- and post-TKA. This more complex design was shared by only 1 of the 5 previously published studies of postural control in persons with TKA.1822

Back to Top | Article Outline


This was the first known study to consider APAs in a population of adults with knee OA and TKA. Other investigators that have considered balance or postural control strategies in this population used measures of balance on an unstable platform,18 quiet sway,19,20 or reactive postural control.21,22 Although there were no changes over time in our participants with TKA, suggesting that the surgery itself had no impact on APAs, we did find between group differences, particularly for the EMG amplitude of the VL and BF during the standing reaching task. The combination of less amplitude of the VL and BF along with the finding of impaired strength in these muscle groups may be the result of impaired neural activation of these muscles, or may be an effort to reduce stress on the involved knee joint. Further investigation is recommended to determine the clinical relevance of the differences in APAs between groups.

Back to Top | Article Outline


The authors thank Lisa Aronson, DPT, Diana Holthaus, DPT, Jill Lindsteadt, DPT, Samantha Minnick, DPT, and Lisa Seigel, DPT for their assistance with data collection; Andrew Coward, MS, Joan Deffeyes, PhD, and Elizabeth Lyden, MS for assistance with data and statistical analysis; and the support of doctoral committee members Judith Burnfield, PT, PhD, Caroline Goulet, PT, PhD, M. Patricia Leuschen, PhD, Wayne Stuberg, PT, PhD, and Ruiping Xia, PhD.

Back to Top | Article Outline


1. Coimbra A, Newton RA. Physical performance measures in known fallers with and without a history of arthritis. Neurol Report. 2002;26:194.
2. Hassan BS, Mockett S, Doherty M. Static postural sway, proprioception, and maximal voluntary quadriceps contraction in patients with knee osteoarthritis and normal control subjects. Ann Rheum Dis. 2001;60(6):612–618.
3. Hinman RS, Bennell KL, Metcalf BR, Crossley KM. Balance impairments in individuals with symptomatic knee osteoarthritis: a comparison with matched controls using clinical tests. Rheumatol (Oxford). 2002;41(12):1388–1394.
4. Masui T, Hasegawa Y, Yamaguchi J, Kanoh T, Ishiguro N, Suzuki S. Increasing postural sway in rural-community-dwelling elderly persons with knee osteoarthritis. J Orthop Sci. 2006;11(4):353–358.
5. Wegener L, Kisner C, Nichols D. Static and dynamic balance responses in persons with bilateral knee osteoarthritis. J Orthop Sports Phys Ther. 1997;25(1):13–18.
6. Singh JA, Vessely MB, Harmsen WS, et al. A population-based study of trends in the use of total hip and total knee arthroplasty, 1969–2008. Mayo Clin Proc. 2010;85(10):898–904.
7. Berth A, Urbach D, Awiszus F. Improvement of voluntary quadriceps muscle activation after total knee arthroplasty. Arch Phys Med Rehabil. 2002;83(10):1432–1436.
8. Mizner RL, Stevens JE, Snyder-Mackler L. Voluntary activation and decreased force production of the quadriceps femoris muscle after total knee arthroplasty. Phys Ther. 2003;83(4):359–365.
9. Mizner RL, Petterson SC, Snyder-Mackler L. Quadriceps strength and the time course of functional recovery after total knee arthroplasty. J Orthop Sports Phys Ther. 2005;35(7):424–436.
10. Pijnappels M, Bobbert MF, van Dieen JH. Control of support limb muscles in recovery after tripping in young and older subjects. Exp Brain Res. 2005;160(3):326–333.
11. Messier SP, Glasser JL, Ettinger WH Jr, Craven TE, Miller ME. Declines in strength and balance in older adults with chronic knee pain: a 30-month longitudinal, observational study. Arthritis Rheum. 2002;47(2):141–148.
12. Pap G, Meyer M, Weiler HT, Machner A, Awiszus F. Proprioception after total knee arthroplasty: a comparison with clinical outcome. Acta Orthop Scand. 2000;71(2):153–159.
13. Wada M, Kawahara H, Shimada S, Miyazaki T, Baba H. Joint proprioception before and after total knee arthroplasty. Clin Orthop Relat Res. 2002(403):161–167.
14. Finch E, Walsh M, Thomas SG, Woodhouse LJ. Functional ability perceived by individuals following total knee arthroplasty compared to age-matched individuals without knee disability. J Orthop Sports Phys Ther. 1998;27(4):255–263.
15. Rossi MD, Hasson S, Kohia M, Pineda E, Bryan W. Mobility and perceived function after total knee arthroplasty. J Arthroplasty. 2006;21(1):6–12.
16. Walsh M, Woodhouse LJ, Thomas SG, Finch E. Physical impairments and functional limitations: a comparison of individuals 1 year after total knee arthroplasty with control subjects. Phys Ther. 1998;78(3):248–258.
17. Swinkels A, Newman JH, Allain TJ. A prospective observational study of falling before and after knee replacement surgery. Age Ageing. 2009;38(2):175–181.
18. Swanik CB, Lephart SM, Rubash HE. Proprioception, kinesthesia, and balance after total knee arthroplasty with cruciate-retaining and posterior stabilized prostheses. J Bone Joint Surg Am. 2004;86A:328–334.
19. Gauchard GC, Vancon G, Meyer P, Mainard D, Perrin PP. On the role of knee joint in balance control and postural strategies: effects of total knee replacement in elderly subjects with knee osteoarthritis. Gait Posture. 2010;32(2):155–160.
20. McChesney JW, Woollacott MH. The effect of age-related declines in proprioception and total knee replacement on postural control. J Gerontol A Biol Sci Med Sci. 2000;55(11):M658–M666.
21. Gage WH, Frank JS, Prentice SD, Stevenson P. Organization of postural responses following a rotational support surface perturbation, after TKA: Sagittal plane rotations. Gait Posture. 2007;25(1):112–120.
22. Gage WH, Frank JS, Prentice SD, Stevenson P. Postural responses following a rotational support surface perturbation, following knee joint replacement: Frontal plane rotations. Gait Posture. 2008;27(2):286–293.
23. Aruin AS, Latash ML. Directional specificity of postural muscles in feed-forward postural reactions during fast voluntary arm movements. Exp Brain Res. 1995;103(2):323–332.
24. Mochizuki G, Ivanova TD, Garland SJ. Postural muscle activity during bilateral and unilateral arm movements at different speeds. Exp Brain Res. 2004;155(3):352–361.
25. Karst GM, Venema DM, Roehrs TG, Tyler AE. Center of pressure measures during standing tasks in minimally impaired persons with multiple sclerosis. J Neurol Phys Ther. 2005;29(4):170–180.
26. Tyler A, Roehrs T, Culcea E, Johnson G, Karst G. Muscle activity patterns of individuals with multiple sclerosis during reaching movements performed while standing. 2002 Abstract Viewer/Itinerary Planner. Washington, DC: Society for Neuroscience. Online. 2002; Program No. 666.15.
27. Bleuse S, Cassim F, Blatt JL, et al. Anticipatory postural adjustments associated with arm movement in Parkinson's disease: a biomechanical analysis. J Neurol Neurosurg Psychiatry. 2008;79(8):881–887.
28. Garland SJ, Stevenson TJ, Ivanova T. Postural responses to unilateral arm perturbation in young, elderly, and hemiplegic subjects. Arch Phys Med Rehabil. 1997;78(10):1072–1077.
29. Horak FB, Esselman P, Anderson ME, Lynch MK. The effects of movement velocity, mass displaced, and task certainty on associated postural adjustments made by normal and hemiplegic individuals. J Neurol Neurosurg Psychiatry. 1984;47(9):1020–1028.
30. Nashner LM, Peters JF. Dynamic posturography in the diagnosis and management of dizziness and balance disorders. Neurol Clin. 1990;8(2):331–349.
31. Equitest System Operator's Manual, Version 8.1. Clackamas, OR: NeuroCom International, Inc; 2003.
32. Ford-Smith CD, Wyman JF, Elswick RK Jr, Fernandez T, Newton RA. Test-retest reliability of the sensory organization test in noninstitutionalized older adults. Arch Phys Med Rehabil. 1995;76(1):77–81.
33. Di Fabio RP. Sensitivity and specificity of platform posturography for identifying patients with vestibular dysfunction. Phys Ther. 1995;75(4):290–305.
34. Tyler AE, Karst GM. Postural behavior of young and healthy elderly during reaching movements. Soc Neurosci Abstr. 1999;25:109.
35. Tyler AE, Karst GM. Timing of muscle activity during reaching while standing: Systematic changes with target distance. Gait Posture. 2004;20(2):126–133.
36. Zipp P. Recommendations for the standardization of lead positions in surface electromyography. Eur J Appl Physiol. 1982;50:41–54.
37. Venema DM, Karst GM. Reliability of electromyography amplitude across sessions. Accessed March 5, 2011.
38. Byrne JM, Gage WH, Prentice SD. Bilateral lower limb strategies used during a step-up task in individuals who have undergone unilateral total knee arthroplasty. Clin Biomech (Bristol, Avon). 2002;17(8):580–585.
39. Mizner RL, Snyder-Mackler L. Altered loading during walking and sit-to-stand is affected by quadriceps weakness after total knee arthroplasty. J Orthop Res. 2005;23(5):1083–1090.
40. Enoka RM. Movement forces. In: Neuromechanics of Human Movement. Third ed. Champaign, IL: Human Kinetics; 2002:57–118.
41. Shield A, Zhou S. Assessing voluntary muscle activation with the twitch interpolation technique. Sports Med. 2004;34(4):253–267.
42. Hodges P, Cresswell A, Thorstensson A. Preparatory trunk motion accompanies rapid upper limb movement. Exp Brain Res. 1999;124(1):69–79.
43. van der Fits IB, Klip AW, van Eykern LA, Hadders-Algra M. Postural adjustments accompanying fast pointing movements in standing, sitting and lying adults. Exp Brain Res. 1998;120(2):202–216.
44. Di Fabio RP. Reliability of computerized surface electromyography for determining the onset of muscle activity. Phys Ther. 1987;67(1):43–48.
45. Hodges PW, Bui BH. A comparison of computer-based methods for the determination of onset of muscle contraction using electromyography. Electroencephalogr Clin Neurophysiol. 1996;101(6):511–519.
Back to Top | Article Outline


MP100 Workstation hardware and AcqKnowledge version 3.8.1 software (BIOPAC Systems, Inc, Santa Barbara, California) was used to collect, store, and view data in real time for the reaching task, gait, and isometric strength testing. All signals were sampled at 1000 Hz using a 16-bit analog-to-digital converter. Computerized algorithms for analysis of data from the reaching task, gait, and isometric strength testing were written in MATLAB® (The MathWorks, Natick, Massachusetts).

Back to Top | Article Outline


An audible tone was the stimulus for participants to initiate the reaching movement. Individuals reached with both arms in the sagittal plane to touch a target that consisted of carbon rubber electrodes mounted on an adjustable lightweight frame. Participants wore a metal thimble on one index finger. Touching the thimble to the target closed a low-voltage circuit to indicate target acquisition. To minimize variation in the reaching activity, participants were instructed to perform the reaching movements as fast as possible.29 Participants were instructed to hold the final reach position for each repetition until a 3-second data collection period was complete.

The accelerometer was attached to the radial side of the wrist, with one of its axes oriented in the sagittal plane and perpendicular to the long axis of the forearm. The signal from this axis was used to identify arm movement onsets for the reaching movement. A computerized algorithm identified the point in time at which the acceleration trace initially exceeded the resting baseline by one standard deviation. Baseline of the acceleration trace was defined as the mean value of the signal over the first 300 ms of the sampling period, which had started 500 ms prior to the tone. Each computer-identified onset was visually confirmed by one investigator. Trials were rejected if the onset was uncertain because of artifacts that may have been due to movement of the hand prior to the reach. Of a total of 600 reaching trials, only 14 (2%) were rejected because of difficulty in identifying a clear onset of wrist acceleration.

Electromyography data were rectified and smoothed using root-mean-square processing with a time constant of 5 ms, and filtered using a second-order 100 Hz low pass Butterworth filter. For each reach distance, the 5 reaching trials were aligned according to the onset of arm movement. Electromyography data were averaged for these trials and onsets of EMG activity were identified from the averaged files using a computerized algorithm. The onset was defined as the point in time when EMG data first exceeded the resting baseline (mean of the first 300 ms of the sampling period) by 2 standard deviations for a period of 30 ms. Each computer-identified onset was visually inspected and adjusted if necessary by one investigator. Using both computerized algorithms and visual inspection is a commonly accepted method to identify EMG onsets.35,4245 Of the 480 samples of EMG data used in the final analysis, 90 (19%) were adjusted from the original computer-identified onset. Twenty-six samples (5%) were rejected due to an inability to identify a clear onset. Electromyography onsets were expressed relative to the time of arm movement onset.

Once the EMG onset was determined for each muscle, the mean EMG amplitude for a 500 ms window beginning at the onset of muscle activity was calculated. The initial burst of activity near the initiation of arm movement was of interest as this is when the central nervous system has to account for destabilization caused by the arm movement. The 500 ms window was chosen after visual inspection of the data indicated that it appeared to best capture the bursts of EMG activity associated with the rapid arm movement. Mean EMG amplitude for each muscle was normalized with reference EMG amplitudes collected during gait.

In addition to the EMG data, COPy data were averaged for the reaching trials to each distance and examined for displacement in the sagittal plane. Both the maximum total COPy displacement, as well as the net displacement (the difference of COPy position before and after the reach) was considered. COPy displacements were normalized for each subject by dividing these values by the subject's foot length.

Back to Top | Article Outline


Participants walked 6 times at a self-selected pace across a 4-meter walkway, with additional distance at each end allowing for acceleration and deceleration. Pulsed infrared beam sensors (Radio Shack, Fort Worth, Texas) at each end of the 4 m measured walking time to allow calculation of each participant's average pace. Trials that fell outside of each participant's average pace (determined at the first session) by ± 10% were repeated.

Before walking, normally open pressure switches (model NO-1R, Tapeswitch, Farmingdale, New York) were applied to the heels of each participant's shoes, with the position of the switches constant between sessions. These switches were used to identify heel strike by closing a 5-volt circuit when pressure was applied. Signals from the heel switches were used to identify each stride. For each muscle, a 100 ms moving window identified the portion of each stride that contained the highest mean EMG amplitude. Mean amplitudes for 12 strides were averaged and served as the reference value for normalization of the EMG data from the reaching task.

Back to Top | Article Outline


Participants were seated upright with back support, hips in 90° of flexion, the knee in 60° of flexion, and the feet unsupported. For stabilization, a Velcro strap was placed around the distal thigh, and participants used handgrips along the side of the seat. A stationary padded bar was placed just proximal to the malleoli, and a strain gauge force transducer was attached perpendicular to the stationary bar. The position of participants and equipment was constant between sessions.

Participants were instructed to slowly build up the amount of torque exerted against the padded bar. Torque data were collected for 3 seconds once participants indicated they were at their maximal effort. Three repetitions were completed for each muscle group. A 1000 ms moving window identified the portion of each of the 3 trials with the highest mean torque for both knee flexion and extension. The greatest mean torque of these 3 trials was normalized to participants' body weight. Cited Here...

anticipatory postural adjustments; balance; electromyography; total knee arthroplasty

© 2012 Academy of Geriatric Physical Therapy, APTA