Reliability of measurements of body center-of-mass momentum during sit-to-stand in healthy adults.

One of the most commonly executed functional activities is the sit-to-stand (STS) transfer. The ability to perform the STS transfer is a prerequisite for upright mobility and is commonly altered in individuals with a variety of neurologic and orthopedic problems, as well as in frail older adult populations. [1-4] Evaluation of STS transfer for assessing postural and movement function may be useful because the task requires a transition from a relatively large, stable base of support in sitting to a considerably smaller base of support in standing. Measurement of physical function often relies on observation and some type of functional scale. [5] Such approaches approaches, however, may be inadequate to characterize neuromusculoskeletal mechanisms[6] that underlie the production of movements such as STS transfers. Recent advances in the movement sciences have provided tools for the development of analytic methods, assessment procedures, and treatment interventions that can be used by physical therapists.[7,8]

Analysis of STS transfers using a biomechanical approach provides information regarding one of the key requirements for the STS transfer, which is to maintain the body's center of mass (CM) within the base of support at die termination of the transfer of weight. Such information pertaining to the control of equilibrium may provide physical therapists with data they can use to guide assessment and treatment protocols. Before clinical methods and procedures can be effectively developed, we contend that measurements that reflect underlying control processes must be shown to be valid and reliable in order to minimize misleading interpretations.

The Study of Movement Performance in Healthy Subjects

The population in which a functional task is investigated is of considerable importance. We contend that studying healthy subjects provides a basis for identifying normative aspects of functional movement performance. The choice of dependent variables, however, dictates the utility of the normative assessment information. Using gait as an example, some would argue normal function may be used as a model against which pathological gait is judged.[9] The assessment of normal and pathological gait through biomechanical measurements that link the cause (kinetics) with the effect (kinematics) is critical if such an assessment is to be useful in identifying alterations m movement performance.[10] A similar rationale holds for the assessment of normative performance using biomechanical variables reflecting balance requirements of STS transfers. Knowledge of the performance characteristics of healthy subjects may serve to improve interpretation of these measurements when pathology exists.[11]

We believe that quantifying the reliability of the biomechanical measurements taken during an STS transfer in a healthy population is necessary in order to interpret information about the variability of the mechanical eve that produce the movement. We argue that by identifying how a healthy population performs STS transfers using reliable and valid output measurements, clinicians can begin to understand how movement function may be altered in patients with disorders of the neuromusculoskeletal system. Thus, the overall advantages of extrapolating quantitative kinetic and kinematic information obtained from research using healthy adults are twofold: (1) potential minimization of inaccurate interpretations of underlying processes governing the control of human movement and (2) a specific understanding of how the neuromusculoskeletal system may change with pathology, injury, and normal development. Such data have the potential to provide a foundation in which to answer clinical questions pertaining to how the movement control system changes for specific cases, how to look for alterations in performance, what changes to look for during the analysis of movement, and most importantly how outcomes of interventions may be quantified.

By establishing the reliability of kinetic and kinematic measurements in healthy adults during STS transfers, normative guidelines for the potential use of measurements of variability as indexes of efficacy of treatment may be formulated. In addition to improvements in mean performance outcome (eg, speed, strength), an improvements consistency of measurements of outcome may be a useful index of treatment efficacy.

Control of the Body's Center-of-Mass Momentum

Some studies[12,13] have quantified the STS movement using kinematic approaches focusing on the assessment of joint angular displacements. Measurements of the excursion of joints and body segments in healthy adults may be valuable as a normative reference for the STS movement.[13] Such approaches, however, do not provide information about where important accelerations are occurring in the system during the task, and they offer limited insight on the underlying control processes governing the STS transfer.

An impulse-momentum relationship has been established for the STS transfer.[14,15] A noncyclical motion such as the STS movement requires a propulsive impulse (eg, the net effect of force acting over a period of time) acting on the body via ground reaction forces (GRFs) in the horizontal direction to initiate forward momentum (ie, the product of body mass and velocity) of the body's CM. This propulsive impulse must then be reversed in direction to become a braking impulse in order to stop the body's momentum, or a forward falling will tend to occur (Figure).[14]

A similar impulse-momentum relationship exists in the vertical direction. Directionally specific differences[14] between horizontal and vertical motions of the CM, however, include a direct relationship between the extent of the vertical motion and the height of the subject and a greater influence of the body's weight (its force vector acting vertically downward) in braking the vertical motion. In contrast to motion in the vertical direction, motion in the horizontal direction is related to standing equilibrium, because the CM must be controlled within rather restrictive limits compatible with the base of support in the anteroposterior direction at the termination of the weight transfer.[14]

There are two advantages in utilizing an impulse-momentum approach when analyzing the STS transfer. First, this approach captures the conditions under which balance control can be investigated. The physical constraints associated with controlling the CM momentum, that is, increasing momentum to perform the voluntary act and subsequently limiting or braking the momentum, are necessary for the control of balance.[16,17] Second, if the dynamics of the movement can be conveyed by using velocity (momentum is the mass product of velocity), it would represent a more accurate approach than using the derivative acceleration because the error inherent to motion analysis would be magnified 10 times for each derivation.18

A valid assessment approach must be established in order to gain insight into underlying control processes governing weight transfer and balance function. The impulse-momentum approach is a means of analyzing the control of the body's CM during a noncyclical action such as STS transfer. Biomechanical laws that govern the requirement of keeping the body's CM within the base of support at the end of a weight transfer justify its use. Furthermore, the impulse-momentum relationship provides a means by which to account for both the kinetics and the kinematics of the action. Thus, the impulse-momentum approach directly quantifies the dynamics (ie, the kinetics or force-time integral and the kinematics or product of mass and velocity of the body's CM) of this particular movement function.

Previous investigations of the impulse-momentum relationships during STS transfers[14,15] have offered insight into the variables that reflect the dynamics of this task. Investigations into the variability of these data, however, are lacking. The purpose of this study was to determine the reliability of measurements of impulse-momentum events underlying the weight transfer of the body's CM in the horizontal and vertical directions during STS transfers.

Method

Subjects

The subjects for this study were 19 healthy adult volunteers (10 female, 9 male) who gave informed consent. The female subjects had a mean age of 31.1 years (SD=3.1, range=27-36), a mean body mass of 60.3 kg (SD=8.5, range=47-74.5), and a mean height of 1.73 m (SD=0.07, range=1.63-1.83). The male subjects had a mean age of 32.4 years (SD=5.2, range=25-38), a mean body mass of 70.9 kg (SD=9.6, range=62.7-85), and a mean height of 1.78 m (SD=0.07, range=1.70-1.88). M6 subject had a history of any major neurological, musculoskeletal, or other medical disorders.

Procedure

Subjects sat comfortably on a chair with a seat height of 0.45 m. This height was chosen as representative of a standard-height chair.[12,19] Upper-limb position was standardized by having the subjects fold their arms in front of their body during the task. This initial posture was adopted to minimize potential intrasubject and intersubject differences in arm motion that might influence the variability of outcome measures. The subjects' feet were placed on the floor in front of the chair. Subjects performed the STS transfer task in blocks of five trials at three different self selected speeds: "as fast as possible," natural speed, and "as slowly as possible."

Instrumentation

A WATSMART motion analysis system(*) was used to record the instantaneous locations of joint centers during the STS transfer. Infrared light-emitting diodes (ILEDs) were taped onto the skin overlying the head of the left fifth metatarsal, lateral malleolus, femoral condyle, greater trochanter, and head of the humerus, facing two infrared cameras located 3.5 m from the center of the force platforms. The cameras were positioned 2 m apart and 2.2 m above the floor. Two strain-gauge force platforms[dagger] were used. One force platform was positioned beneath the chair to measure the time when each subject lost contact with the chair seat, and the other force platform was positioned under the subject's feet to measure the propulsive and braking impulses acting on the subject in the horizontal and vertical directions.

An IBM-AT compatible computer[double dagger] was used for the motion analysis data collection and was synchronized with a PDP 11/73 computer,[section] which was used to collect the force platform data. Data were sampled at a rate of 100 Hz for 5 seconds. Recently, the reliability and validity of measurements obtained with the WATSMART system have been investigated.[20] Prior to each experimental session of the present study, the motion analysis system was calibrated using a calibration frame with REDS placed m known positions. A drift test was performed for each force platform, and an offset reading was taken. The accuracy and the validity of measurements obtained by the motion analysis system and force platforms were evaluated by a series of tests. The estimated measurement error was judged to be acceptable in this study and in a previous study[14] according to the following rationale. In one test, the resultant force acting on a subject during the performance of die M transfer was computed from the acceleration of the CM derived from the motion analysis and from the GRF derived from the two force platforms. The mean for the root-mean-square of the differences between the two results for all the trials analyzed was 18 N (SD=2), which was less than 8% of the peak value measured in this task.

Data Processing

To minimize inherent to the environment, the joint location coordinate data derived from the anatomical markers were smoothed using a low-pass second-order Butterworth digital filter with a cutoff frequency of 3 Hz. The instantaneous location of the CM was computed based on bodysegment parameters.[21] Central finite difference methods were applied to compute horizontal and vertical components of the velocity of the CM. The corresponding linear momentum was then computed.

Five events were identified for each trial in order to operationally define the dependent variables. The time when the magnitude of the horizontal momentum just exceeded 7% of its peak value represented the beginning of the STS transfer, and the end of the STS transfer was identified when the magnitude of the horizontal momentum fell just below 70/o of its peak value.[14] The criterion level for task initiation and termination was chosen to distinguish the intended act of STS transfer from postural sway encountered during stationary sitting and standing. Peak horizontal momentum and peak vertical momentum were also identified. The time when the momentum achieves peak values also represents the time when the propulsive impulses reverse directions to become braking impulses.[14] The time when the subject lost contact with the chair was determined by the vertical force from the force platform beneath the chair.

Data Analysis

The means, standard deviations, and ranges of the dependent variables between subjects over all trials were obtained. Temporal variables (measured in seconds) included the time to peak horizontal momentum and peak vertical momentum and the time when the subject lost contact with the chair. Magnitude variables (measured in kilogram-meters per second) included the peak horizontal momentum and the peak vertical momentum. Subject x trial two-way analyses of variance (ANOVAs) were performed using the SYSTAT statistical package[parallel] for the five dependent variables. Separate ICCs were performed to determine the reliability of the data. The ICC (2,1) as described by Shrout and Fleiss[22] was chosen because trials were considered random effects. Findings are thus generalizable to other trials for a particular subject.

Results

All subjects performed the STS transfer at the fast speed. Eighteen of the 19 subjects performed the STS transfer at a natural movement speed. Slow-speed magnitude data are based on 19 subjects, and slow-speed temporal data are based on 18 subjects due to missing data values attributable to inadvertent coverage of one or more diodes during data collection.

Three trials from three separate subjects for peak vertical momentum at the fast speed were excluded from the data analysis. Using suggested values[23] adopted a priori, these values were identified as statistical outliers, as they exceeded 3 1/2 standard deviations from each subject's remaining mean (trial) values. Careful examination of the data collection and processing procedures revealed no error in measurement or recording, equipment malfunction, or miscalculation.[24] A potential procedural factor that may have accounted for these values, however, was identified. In each of the three trials, the outlier values occurred in the first recorded trial of STS transfer for the fast movement condition. These values were always significantly lower than those of the remaining four trials. We believe an inconsistency in either the instruction to move as fast as possible or the subjects' understanding of the instruction within the first trial resulted in these outlier values. Highly consistent peak vertical momentum values for the remaining trials suggests that this inconsistency was resolved by the second trial. Such extreme outlier values were not observed for these or other subjects for any other measured variables or movement speeds.

The group mean duration of the performance of STS transfer was 1.39 seconds (SD=0.26) for fast movements, 1.85 seconds (SD=0.28) for natural movements, and 2.94 seconds (SD=0.51) for slow movements. The group means, standard deviations, and ranges for all temporal and magnitude variables across trials are summarized in Table 1.

Table 1. Group Means, Standard Deviations, and Ranges for Horizontal-and
Vertical-Momentum Temporal and Magnitude Variables During Sit-to-Stand Transfer
Speed       THM(a)   TVM(b)   OFF(c)   PHM(d)    PVM(e)
Fast(f)
  X         0.40     0.62     0.47     38.02      68.15
  SD        0.12     0.13     0.12      9.00      17.88
  Range     0.24     0.43     0.29     24.32      34.60
            0.89     1.07     0.97     65.27     115.86
Natural(g)
  X         0.54     0.89     0.63     34.26      46.38
  SD        0.10     0.15     0.12      6.60      10.86
  Range     0.40     0.62     0.45     24.00      27.54
            0.87     1.43     0.93     52.77      73.29
Slow
  X         0.85(f)  1.39(f)  0.97(f)  24.51(g)   25.28(g)
  SD        0.22     0.39     0.23      6.38       8.48
  Range     0.42     0.61     0.53     15.58      10.35
            1.50     2.18     1.57     44.34      47.84
(a) THM = time (in seconds) to peak horizontal momentum.
(b) TVM = time (in seconds) to peak vertical momentum.
(c) OFF = time (in seconds) to when the subject lost contact with the chair.
(d) PHM = peak horizontal linear momentum (in kilogram-meters per second).
(e) PVM = peak vertical linear momentum (in kilogram-meters per second).
(f) N = 19.
(g) N = 18.

When computing the ANOVAs, the between-subject effect was significant (P<.01) for all variables and there was a nonsignificant trial X subject interaction (subjects performed consistently across trials) (P>.05) for all variables except time to peak vertical momentum at the slow movement (Tab. 2). A significant between-subject effect and nonsignificant trial X subject interaction are necessary prerequisites for the ICC to be useful in representing reliability.[25]

[TABULAR DATA OMITTED]

Temporal Measures

At the fastest STS transfer speed, ICC values were low and ranged from .28 to .33. The ICC values at the natural speed were moderate (ICC=.51-.64), whereas reliability coefficients at the slow speed were moderately high (ICC=.72-.75). The ICC values for an temporal measures are presented in Table 3.

Magnitude Measures

Peak horizontal and vertical momentum ICC values were high ([is greater than or equal to]81) for all speeds of movement. The ICC values for all magnitude measurements are shown in Table 3.

Table 3. Intraclass Correlation Coefficient
(2,1) Reliability Estimates of
Horizontal- and Vertical-Momentum
Temporal and Magnitude Variables During
Sit-to-Stand Transfer (P<.01)
           Speed
           Fast(a)  Natural(b)  Slow
Time
  THM(c)   0.30     0.57        0.75(a)
  TVM(d)   0.33     0.51        0.72(a)
  OFF(e)   0.28     0.64        0.74(a)
Magnitude
  PHM(f)   0.85     0.87        0.88(b)
  PVM(g)   0.95     0.86        0.81(b)
(a) N = 19.
(b) N = 18.
(c) THM = time (in seconds) to peak horizontal
momentum.
(d) TVM = time (in seconds) to peak vertical
momentum.
(e) OFF = time (in seconds) to when the subject
lost contact with the chair.
(f) PHM = peak horizontal linear momentum (in
kilogram-meters per second).
(g) PVM = peak vertical linear momentum (in
kilogram-meters per second).

Discussion

The reliability of measurements of the horizontal and vertical momentums of the CM during STS transfers in healthy adult subjects was investigated. The results indicated that magnitude variables are very consistent within and between subjects and demonstrate high reliability regardless of speed of STS transfer. Temporal variables exhibited a range of reliability estimates, which varied with changing speed.

Sources of Measurement Error

Rothstein[26] has identified three potential sources of error contributing to a lack of reliability in measurements: instrumentation error, inherent variability in the subject's performance, and evaluator error. Qualitative observations can be influenced by a rater's experience and assumptions, and functional scales may be weakened by poor validity. These weaknesses may be minimized by the appropriate choice of instrumentation with which to best obtain measurements. The present motion analysis system has been found to yield valid and reliable measurements.[20] Furthermore, the instrumentation used in our study has been evaluated for accuracy and variability (see "Instrumentation" section) and was calibrated prior to each data collection session. All variables measured by the motion analysis system reflect peak values; therefore, sufficient detectability to change (sensitivity) exists in order to capture the nature of these variables. Thus, measurement error due to the instrumentation has been minimized.

Evaluator error was greatly minimized in this study. For example, during movement analysis, the beginning and end of STS transfer were operationally defined (see "Data Processing" section). Furthermore, identification of these events was accomplished by graphical analysis programs. Thus, subjective interpretation of the results by the rater was minimized. Finally, instructions given by the evaluator and placement of the ILEDs were standardized.

Variability in subject performance may be influenced by such factors as seat height in relation to subject height, differing initial postures, individual anthropomorphic differences (eg, body mass and height), and variations in interpreting instructions concerning movement speed. Such variables may influence the way resultant joint torques[27] are generated, particularly with changing seat height.[28] Effects of anthropomorphic differences may become less of a factor, however, when subjects are allowed to adopt a comfortable leg position,[29] as they did in this study.

A standardized chair height was used to approximate daily performance in which seat height is nonvariable in a given environment. It is important to note that relative height of the chair (ratio of seat height and subject's height or lower leg length) will likely be a confounding factor affecting the variability in the outcome of the joint motion and torques,[28] yet this factor should have an equivalent effect on the same subject performing the task across different speeds of ascent.

A standardized initial upper-limb position was chosen to minimize the effects of inter-subject variability during the movement. Although such standardization may compromise the generalizability the measurement to behavior in natural surroundings, failure to incorporate such factors may increase variability and unduly confound interpretation of the results.

In contrast, subjects were allowed to adopt a comfortable and natural foot position. Foot position may have a greater influence on the performance of the task in that it directly alters the subject's base of support. This altered base of support influences the excursion of the body's CM from the initial seated base of support to the end-position base of support in standing. A standardized and potentially unfamiliar initial foot/leg position may place additional demands on trunk motion and possibly alter trunk and thigh muscle activity.[29] A systematic evaluation of these influences as movement speed changes would be useful to understand their effects on the execution of this task.

Reliability of Temporal Variables

Measurements for time to peak horizontal momentum and peak vertical momentum and time to when the subject lost contact with the chair exhibited reliability estimates ranging from low to moderately high. There are two possible explanations for low reliability.[24,25] The first, lack of agreement among judges trials), can be evaluated by examining the trial subject interaction term. In this case, a nonsignificant interaction is required for the ICC to be useful as a reliability estimate.[25] that is, the spread of the intrasubject scores (or differences between trials) must be at a minimum. No such interaction was found at fast or natural movement speeds. The time to peak vertical momentum during slow STS transfer, however, exhibited a significant interaction. Therefore, though the ICC was moderately high, interpretation of this reliability estimate must be made with caution, as subjects did not demonstrate sufficient repeatability with respect to this variable for a slow STS transfer.

The standard deviations, which reflect the absolute variability of the entire sample, were twice as large for slow speeds as compared with natural and fast movement speeds. These standard deviations are representative of the within-subject standard deviations in that in all except one subject, the standard deviations for slow-movement temporal variables were always greater than those values at natural and fast movement speeds. Newell and colleagues[30] examined movement temporal variability across a range of movement times and velocities during a linear timing task. Within-subject standard deviation of movement time was found to increase as the velocity of movement became slower.

The second possible explanation for low reliability--lack of variability among subjects--can be evaluated by examining the between-subject (main) effect. Homogeneous subject scores would result in a nonsignificant source of variance.[24] in our study, all between-subject effects were significant across all speeds.

As subjects moved more slowly, the reliability estimates increased such that, for slow movements, reliability values were moderately high. Although no definitive explanation is readily apparent, the nature of the movement performance across speeds may account, in part, for this result. For example, durations of the STS transfer were somewhat closer for natural and fast movement speeds (within 0.5 seconds). Subjects took, on average, I second longer to perform a slow STS transfer versus natural and fast transfers. The only constraints on movement speed in this study were through verbal instructions. Thus, temporal accuracy was not a formal requirement. Because subjects were not trained to perform the STS transfer at a constrained movement time, a greater range of reliability estimates may exist for temporal measurements. If such constraints were imposed, greater temporal separation between speeds and less overlap in the ranges for the temporal variables would exist (see event durations in the "Results" section and Tab. 1). if subjects became well practiced at each specific speed, the ICCs might more closely approach higher values. Finally, it would appear that a successful STS transfer may be more dependent on the temporal sequencing of the peak horizontal momentum, lift-off from the chair, and peak vertical momentum with respect to each other[14] than the repeatability of time to peak (or time to lift-off) measurements alone.

Reliability of Magnitude Variables

Measurements of the peak horizontal and vertical momentums exhibited the highest reliability estimates. These high ICCs indicate that these measurements are consistent and reproducible for healthy adults across a range of movement speeds. Interestingly, peak vertical momentum magnitudes progressively increased as speed of ascent increased, whereas peak horizontal momentum magnitudes exhibited disproportionately smaller increases from slow to natural to fast movement speeds (Tab. 1). As the weight of the body opposes vertical motion, an increasing propulsive impulse in the vertical direction would likely be less destabilizing to equilibrium than one in the horizontal direction. The more modest changes in peak horizontal momentum across movement speeds along with the high ICC values suggest that motion of the CM in the horizontal direction of the sagittal plane may be a relatively stable feature of the STS movement, particularly within a range of natural to fast self-selected speeds. Such restrictions in generating horizontal momentum of the CM may be important for maintaining balance in the forward direction.[14]

Similarities exist between the characteristics horizontal sagittal-plane momentum magnitudes and ICCs for the STS transfer task and between the magnitudes and ICCs of propulsive and braking impulses, as revealed by the resultant GRF component in the horizontal frontal plane for standing single-leg flexion movements.[31] Despite changes in the overall speed of movement, both activities demonstrate stable peak momentum profiles across a range of natural to rapid self-selected speeds. In such cases, motion of the body's CM may be actively constrained for task-specific directions of motion that normally represent greater challenges to balance when moving from a larger (seated position, bipedal stance) to a smaller (completion of STS transfer, single-limb stance) base of support.

Clinical Implications

The temporal and magnitude measures raise important issues regarding movement variability across speeds and interpretation of ICC reliability estimates. First, ICC values and the variability of the data set should both be considered when interpreting results from research reports pertaining to reliability of measurements. Second, die consistency of temporal events can vary at different speeds at which a movement is assessed. Third, it would appear that the consistency of measurements of certain temporal aspects of movement differ from magnitude variables over a range of speeds.[31]

Clinical implications arise based on this information. First, when clinicians evaluate functional movement performance over a range of speeds, they should recognize that the reliability of certain measurements may not be the same across speeds, despite the fact that the overall goal (ie, to stand) has not changed. Second, if similar repeatability of measurements exists for patient populations, then utilizing functional biomechanical measures based on magnitude variables over temporal variables may be warranted. For example, the incorporation of force platforms can provide data on the magnitude of the propulsive and braking impulses generating motion of the body's CM during a functional movement.

The impulse-momentum principle provides a means of identifying and investigating direct and quantifiable variables underlying weight transfer and equilibrium control during STS transfers, and we believe is applicable to a variety of movement tasks commonly used in the clinical assessment of posture and balance. Although measurement of STS transfer performance using computerized kinematic analysis is not, at present, readily available to clinicians due to the cost and technical nature of the instrumentation, such an approach could provide the clinician with an important foundation with which to identify key elements in this and other tasks and with information on how these elements may normally vary according to changes in speed. At the present time, however, data pertaining to the GRFs (impulses) generating motion of the body's CM are more easily gathered through the use of force platforms. Reliability of quantitative data gathered from such instrumentation will be important if everyday tasks such as STS transfers are to be useful as part of future more global functional assessment procedures in patient populations.

It is unclear why reliability of the specific temporal variables of horizontal and vertical momentum measurements investigated in this study did not compare with reliability of magnitude measurements. Furthermore, how die temporal and magnitude variables will compare with those gathered in patient populations remains to be established. Measurements that directly reflect the CM momentum in directions that govern the destination of the projected location of the CM with respect to the base of support at the completion of a weight transfer may provide valuable information in those patients whose balance may be compromised.[17] Such measurements afford the potential to be valid as well as reliable indicators of impaired balance function. Further research using patient populations is currently under way.

Conclusions

The reliability of measurements of motion of the body's CM during STS transfers using an impulse-momentum approach has been investigated. Temporal variables exhibited a range of variability and reliability estimates, dependent on the speed of STS transfer. Measurements of vertical and horizontal magnitude variables were highly reliable across a range of speeds, suggesting that they are very consistent and reproducible in healthy adults.

In light of these results, such an approach could now be useful in quantifying alterations m movement performance in individuals with disorders of the neuromusculoskeletal system if they prove to be valid indicators for prognostic or diagnostic purposes. Furthermore, this approach provides an important foundation for understanding the biomechanics of the STS transfer and other commonly evaluated functional tasks. It has the potential for providing useful data for advancing assessment and treatment procedures in the clinical setting.

Acknowledgment

We thank Karen W Hayes, PhD, PT, for her statistical assistance. [dagger]Advanced Medical Technology Inc, 141 California St, Newton, MA 02158 [double dagger]PC's Limited 286, Dell Computer Corp, 1611 Headway Cir, Bldg 3, Austin, TX 78754. [section]Digital Equipment Corp, 146 Main St, Maynard, MA 01754. [parallel]SYSTAT Inc, 1800 Sherman Ave, Evanston, IL 60201.

References

[1] Coghlin SS, McFayden BJ. Transfer strategies used to rise from a chair in normal and low back pain subjects. Clin Biomech, 1994;9: 85-92. [2] Yoshida K, Iwakura H, Inoue F. Motion analysis in the movements of standing up from and sitting down on a chair. Scand J Rehabil Med. 1983;15:133-140. [3] Engardt M, Ribbe T, Olsson E. Vertical ground reaction force feedback to enhance stroke patient's symmetrical body-weight distribution while rising/sitting down. Scand J Rehabil Med. 1993;25:41-48. [4] Alexander NB, Schultz AB, Warwick DN. Rising from a chair: effects of age and functional ability on performance biomechanics. J Gerontol. 1991;46(3):M91-M98. [5] Jette AM. State of the art in functional status assessment. In: Rothstein JM, ed. Measurement in Physical Therapy. New York, NY: Churchill Livingstone Inc; 1985;7:137-168. [6] Lee WA. A control systems framework for understanding normal and abnormal posture. Am J Occup Ther. 1989;43:291-301. [7] Rogers MW. Motor control problems in Parkinson's disease. In: Contemporary Management of Motor Control Problems. Proceedings of the II Step Conference. Alexandria, Va: Foundation for Physical Therapy Inc; 1991: 195-208. [8] Carr JH, Shepherd RB. A motor learning model for rehabilitation. In: Carr JH, Shepherd RB, eds. Movement Science: Foundations for Physical Therapy in Rehabilitation. Rockville, Md: Aspen Publishers Inc; 1987:31-36. [9] Perry J. Gait Analysis.- Normal and Pathological Function. Thorofare, NJ: SLACK Inc; 1992:xv-xvii. [10] Winter DA. Concerning the scientific basis for the diagnosis of pathological gait and for rehabilitation protocols. Physiotherapy Canada. 1985;37:245-252. [11] Nawoczenski DA, Sharp WB, Maiers DJ, et al. Reliability of performance measurements obtained using the Stability Testing and Rehabilitation Station (STARStation). Phys Ther. 1991;71:706-714. [12] Jeng SF, Schenkman M, Riley PO, Lin SJ. Reliability of a clinical kinematic assessment of the sit-to-stand movement. Phys Ther. 1990; 70:511-520. [13] Nuzik S, Lamb R, VanSant A, Hirt S. Sit-to-stand movement pattern: a kinematic study. Phys 7her. 1986;66:1708-1713. [14] Pai YC, Rogers MW. Control of body mass transfer as a function of speed of ascent in sit-to-stand. Med Sci Sports Exerc. 1990;22: 378-384. [15] Pai YC, Rogers MW. Segmental contributions to total body momentum in sit-to-stand. Med Sci Sports Exerc. 1991,23:225-230. [16] Pai YC, Naughton BJ, Chang RW, Rogers MW. Control of body center of mass momentum during sit-to-stand among young and elderly adults. Gait & Posture. 1994;2: log-116. [17] Pai YC, Lee WA. Effect of terminal constraint on control of balance during sit-to-stand. Journal of Motor Behavior. 1994;26: 247-256. [18] Wood GA. Data smoothing and differentiation procedures in biomechanics. Exerc Sport Sci Rev. 1982;10:308-361. [19] Wheeler J, Woodward C, Ucovich RL, et al. Rising from a chair: influence of age and chair design. Phys Ther. 1985;65:22-26. [20] Scholz JP. Reliability and validity of the WATSMART(TM) three-dimensional optoelectric motion analysis system. Phys Ther. 1989;69: 679-689. [21] Winter DA. Biomechanics of Human Movement. New York, NY: John Wiley & Sons Inc; 1979:151-152. [22] Shrout PE, Fleiss JL. Intraclass correlation: uses in assessing rater reliability. Psychol Bull. 1979;86:420-428. [23] Stevens J. Applied Multivariate Statistics for the Social Sciences. Hillsdale, NJ: Lawrence Erlbaum Associates Inc; 1986: chap 1. [24] Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. East Norwalk, Conn: Appleton & Lange; 1993: 465-525. [25] Krebs DE. Computer communication. Phys Ther. 1984;64:1581-1589. [26] Rothstein JM. Measurement and clinical practice: theory and application. In: Rothstein JM, ed. Measurement in Physical Therapy. New York, NY: Churchill Livingstone Inc; 1985;7:1-46. [27] Pai YC, Rogers MW. Speed variation and resultant joint torques during sit-to-stand. Arch Phys Med Rehabil. 1991;72:881-885. [28] Rodosky MW, Andriacchi TP, Andersson GBJ. The influence of chair height on lower limb mechanics during rising. J Orthop Res. 1989;7:266-271. [29] Stevens C, Bojsen-Miller F, Soames RW. The influence of initial posture on the sit-to-stand movement. Eur J Appl Physiol. 1989;58: 687-692. [30] Newell KM, Hoshizaki LEF, Carlton MJ, Halbert JA. Movement time and velocity as determinants of movement timing accuracy. Journal of Motor Behavior. 1979; 1 1:49-58. [31] Hanke TA, Rogers MW. Reliability of ground reaction force measurements during dynamic transitions from bipedal to single-limb stance in healthy adults. Phys Ther. 1992; 72:810-816.

TA Hanke, PT, is Instructor in Physical Therapy, Programs in Physical Therapy, Northwestern University Medical School, Chicago, IL 60611.

Y-C Pai, PhD, is Assistant Professor, Programs in Physical Therapy, Northwestern University Medical School, 345 E Superior St, Room 1323, Chicago, IL 60611 (USA). Address all correspondence to Dr Pai.

MW Rogers, PhD, PT, is Associate Professor, Programs in Physical Therapy, Northwestern University Medical School.