| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a Department of Mechanical Engineering, University of Michigan, Ann Arbor
b Department of Biomedical Engineering, University of Michigan, Ann Arbor
c Institute of Gerontology, University of Michigan, Ann Arbor
d Department of Internal Medicine, University of Michigan, Ann Arbor
e Ann Arbor VA Medical Center GRECC, Michigan
James A. Ashton-Miller, University of Michigan, MEAM, GGB 3208, Ann Arbor, MI 48109-2125 E-mail: jaam{at}umich.edu.
Decision Editor: John E. Morley, MB, BCh
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Methods. Twenty young (mean age 25 years) and twenty older (mean age 70 years) volunteers, with equal gender representation, performed a seated arm-movement task under three conditions: Condition 1, in which subjects were instructed to raise the hands upon cue as quickly as possible from thigh level to a shoulder height target; Condition 2, in which subjects were instructed as in Condition 1 with the addition of intercepting a swinging pendulum at the prescribed hand target; and Condition 3, in which subjects were instructed as in Condition 2 but were asked to wait as long as possible before initiating hand movement to intercept the pendulum. Arm movements were quantified using standard kinematic techniques.
Results. Age (p < .01) and gender (p < .05) affected hand movement times. In Conditions 1 and 2, the older women required 20% longer movement times than the other subject groups (335 vs 279 milliseconds; p < .01). In Condition 3, shorter movement times were achieved by young men (20%; p = .002) and older women (10%; p = .056) as compared with their respective performance in Conditions 1 and 2 because they did not fully decelerate their hands. The other groups slowed their movements in Condition 3.
Conclusions. Age, gender, and perceived threat significantly affected movement times. However, even the slowest movement times were well within the time available to deploy the hands in a forward fall to the ground.
THE upper extremities are the first line of defense in protecting the head and body from impact (1)(2). This often involves moving the hands rapidly into a protective posture ready for the collision. Although the kinematics of nonprotective rapid arm movements have received some attention (3)(4), it is not known whether the speed of protective movements is adversely affected by age.
It is known that the maximum speed of a nonprotective movement is inversely related to the terminal accuracy required, as predicted by Fitts' Law (5); that rapid nonprotective movements slow with age (6)(7); and that elderly persons are at higher risk for fall-related injuries than young persons (8)(9). It would seem reasonable that if declines in the capacity of elderly persons to move their hands rapidly in a protective manner were found, then interventions could usefully be targeted at ameliorating such a deficit.
We therefore tested the hypotheses that neither age nor gender would affect the time required to perform a protective maneuver or its maximal velocity and acceleration. Furthermore, we tested the secondary hypotheses that neither age nor gender would affect the percentage of hand movement capacity (velocity and acceleration used, expressed as a percentage of maximum available capacities) utilized for protection in a task where the consequences of failure were perceived as threatening.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Subjects were asked to perform three tasks (no-threat, low-threat, and high-threat) while seated.
No-Threat Condition
Two no-threat trials were conducted. After practice, each subject was instructed to start with his or her hands placed on the anterior surface of the thighs and then, upon an audio cue, raise the hands forward as quickly as possible to a break-away target positioned at approximately shoulder height, with the elbow extended to 120° and the fingers pointing upward (Fig. 1). Subjects were given a countdown after which a manually-triggered audio cue signaled the subjects to raise their hands to within a "few inches" of the targets.
|
High-Threat Condition
Eight additional trials were performed in which the subjects were asked to wait "as long as possible following the cued release of the pendulum" before initiating movement to intercept the oncoming pendulum, thus getting their hands into position "just in time."
The pendulum was constructed of a pair of aluminum tubes (outer diameter, 2.6 cm), which were attached to a single hinge joint at the ceiling and were connected together at the bob. The bob consisted of a mounting block, a load cell, and a 12-cm by 20-cm plywood impactor for each hand. The impactor surface was covered with a layer of 2.4-cm-thick stiff foam padding.
An OPTOTRAK (Northern Digital, Waterloo, Ontario, Canada) system was used to measure arm segment and pendulum kinematic data at 300 Hz from a right lateral view. The kinematic data were low-pass filtered (4th order Butterworth; MATLAB, Natick, MA) with a cut-off frequency of 50 Hz. Pairs of infrared light-emitting diode markers were placed on the upper arm, the forearm, and the impactor head and laterally at the base of the neck and on the back of the chair.
Three measures of human performance were used to analyze the data (Fig. 2): (i) the movement time (MT), which was determined as the time from when the velocity of the wrist first exceeded 0.2 m/s until it decreased below 0.4 m/s or until impact occurred (when the force measured at the pendulum exceeded 10 N); (ii) the maximum linear acceleration; and (iii) the maximum linear velocity reached by the wrist during the motion. For each subject the three measures from the best trial were calculated for every test condition, where "best" was defined as the trial with the shortest movement time. The 5 out of 120 best trials in which the movement distance was less than 2 SD below the group mean were excluded in favor of the next fastest trial.
|
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
In the high-threat condition, the YM displayed significantly shorter MT than all other subject groups (YM: 226 milliseconds; YW: 274 milliseconds; OM: 285 milliseconds; OW: 292 milliseconds) (see Table 1 ). The mean ± SD point-to-point distance moved by the hands was 31 ± 3 cm. Distances were similar for the other test conditions. We observed that the distance was shorter for older subjects than for younger subjects (30 vs 32 cm), although this difference was not significant. MT was not correlated to the start-to-stop straight-line distance for any subject group (R2 < .1).
A reduced usage of capacity could be expected in the high-threat case, if a more cautious strategy employed a submaximal movement velocity, because of the perceived threat associated with missing the desired target. In the no- and low-threat conditions, adjustments could still be made to the final hand positioning after the initial movement.
Capacity usage varied with age and gender. The OM exhibited significantly lower maximal acceleration (39% lower) and peak velocity (22% lower) during the high-threat condition as compared with the low-threat condition (Fig. 3). Likewise, the YW also demonstrated a significantly lower acceleration (30% lower), although the peak velocity was not significantly lower than in the no-threat condition (15%; p = .056). OW and YM used the same peak velocity and acceleration in the high-threat condition as they did in the low-threat condition (Fig. 3). Hence, all of the YM (10 out of 10) and most of the OW (7 out of 10) reduced their MT in the high-threat condition to less than 95% of their time in the low-threat condition, whereas less than half of the YW (4 out of 10) and the OM (2 out of 10) did the same. On average, the OW and YM used less MT in the high-threat condition than in the low- and no-threat conditions (approximately 20% lower), whereas the YW and OM did not do so (Table 1 ). Although this difference was significant only in the YM, it approached significance in the OW as well (p = .056).
|
Three subjects were retested at a later date, and the learning effect from the previous session led to a 10% to 14% improvement in the average MT measured during the second session. One of the three retested subjects had her best time in the first session.
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Strength loss with age has been well documented. For example, upper-extremity strength has been shown to decrease by approximately 20% in both men and women between the ages of 30 and 70 years (13). Between the ages of 20 and 75 years, MT has been found to increase by 50% and 100% in nonprotective movements performed by women (7) and men (6), respectively. In the no- and low-threat conditions, however, we did not observe such a slowing in the OM subjects, but we did find a 24% difference (70 milliseconds) between YW and OW, indicating a moderate reduction in OW capacity. The absence of a pronounced age effect in the no-threat condition could be due to a sampling bias because this sample of older subjects was highly active (14). The modest age effect is consistent with the relatively small age effects on MT found in lower-extremity protective responses (15).
There are relatively few studies of the effect of age on physical risk-taking behavior. This may be one of the first studies to physically quantify the effect of age on risk-taking when an individual is faced with performing a familiar protective task, because risk-taking is more commonly assayed using subjective self-report measures ((16),(17)). Others have found that older adults are more conservative or error-averse in making nonprotective rapid movements (18). In the present high-threat test condition, subjects self-selected the level of risk associated with the task. The most aggressive (risky) strategy required the subjects to use their full capacity and accelerate their hands until the time of impact. All the subjects, however, decelerated to some degree before impact in the high-threat condition, suggesting that not even the YM and OW used this most aggressive strategy.
Movement time could be expected to be shorter in the high-threat condition because it was not necessary to decelerate the hand before intercepting the target. Thus, in the no-threat and low-threat protocols the average velocity was 51% of the peak velocity (19), but in the high-threat case the subjects achieved an average velocity that was 64% of the peak velocity. However, only the YM and the OW were willing to use a strategy aggressive enough to lower their MT from that of the no-threat condition.
YM and OW used the same peak velocity and acceleration in the high-threat protocol as in the no-threat protocol, whereas OM used lower levels of velocity and acceleration under the higher risk conditions and YW used a lower level of acceleration. This suggests that the OM and YW were concerned with the accuracy of their hand positioning when attempting the most time-critical task, apparently choosing slower but more accurate movements.
One limitation of the present study was that we could not be absolutely certain that we measured maximum capacity in the no-threat case. We believe, however, that because neither acceleration nor peak velocity increased when the threat of an approaching pendulum was introduced in the low-threat condition, our capacity measures were reasonable. Another limitation of this study was the lack of experimenter control over the trajectory of the upper extremities in these movement tasks. Whereas this has validity in terms of how activities of daily living are performed, it limits how precisely one can compare movement times. This is because subjects used a variety of trajectories to reach the interception point and did not achieve identical final elbow angles in every trial. The finding that MT was not correlated to the point-to-point straight-line length of the hand trajectory used by the subject might be surprising. However, the explanation for this is that there was a strong correlation between distance and peak velocity, with the subjects who had larger point-to-point distances also having higher peak velocities. Despite the seated posture, the range of hand movement distances was not dissimilar to those used in positioning the arms for impact in a forward fall (20).
Taking larger risks in activities of daily living may put an individual at greater risk for falls. YM and OW demonstrated the highest risk-taking in this study. Risk-taking strategies could also affect fall arrest strategies. For example, using a submaximal movement velocity to achieve greater accuracy could result in the hand arriving too late to acquire a targeted hand support. Furthermore, the data indicate that it is important to consider subject perceptions of the amount of risk involved when assessing a patient's ability to make a rapid protective movement. Our results clearly demonstrate that results obtained from no- and low-threat test conditions could not reliably be used to predict performance in what subjects perceive to be a high-threat condition when the spatial accuracy required remains unchanged. Last, although we did observe significant slowing of MT with age, it would not appear to be large enough to affect the biomechanics of fall arrests with the arms.
|
Acknowledgments |
|---|
Kurt M. DeGoede is currently at Elizabethtown College, Elizabethtown, PA.
Received September 7, 2000
Accepted September 12, 2000
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. N. Robinovitch, S. C. Normandin, P. Stotz, and J. D. Maurer Time Requirement for Young and Elderly Women to Move Into a Position for Breaking a Fall With Outstretched Hands J. Gerontol. A Biol. Sci. Med. Sci., December 1, 2005; 60(12): 1553 - 1557. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
