Journal of Gerontology: MEDICAL SCIENCES
1999, Vol. 54A, No. 12, M629-M634
Copyright 1999 by The Gerontological Society ofAmerica
Tai Chi Practice Reduces Movement Force
Variability for Seniors
JinH.Yan
Departmentof Health and Human Performance, The University of Houston,Texas.
Background. The purpose of this study was to examine whether Tai Chi practice can reduce the inconsistency of arm
movement force output in older adults.
Methods. Twenty seniors took part in the 8-week-Iong exercise intervention program (12 in Tai Chi practice, M = 79.3
years, SD =2.4; and 8 in a locomotor activity group, walking or jogging, M =79.5 years, SD = 1.9). Linear and curvilinear
manual aiming movements were tested at the beginning (pretest), during 4th week (retest), and the end of the exercise program
(post-test). The measure of vertical pressure on the surface of a tablet served as the dependent variable.
Results. The findings suggest that the Tai Chi participants significantly reduce more pressure variability than the participants
in locomotor activity group after 8 weeks of practice. Additionally, seniors produced higher pressure variability in the curvilinear task than in the linear task.
Conclusions. Evidence from this study proposes that Tai Chi practice may serve as a better real world exercise for reducing
force variability in older adults' manual performance.
DESEARCHERS typically use the task of aiming arm move-
.l'.. ment to study senior adults' motor control (1-3). Senior
adults' arm movements are normally slower, more variant, and
less smooth than their younger counterparts (3-5). In terms of
generating force for movement execution, senior adults often
demonstrate greater variability than young adults (6-8). The information from measuring force variability, described as the
fluctuation of force output during movement implementation
(9), is indicative of the consistency of motor performance (10).
Several reasons may explain why movements become slower
and force variability increases as individuals become older. One
such explanation is that the capability of integrating sensorimotor information relative to force production declines for senior
adults, which accounts for the decreased movement speed and
the increased variability in isometric force (11-12). Recently,
Walker and colleagues (13) suggested that movement slowness
of senior adults may be related to "noise" in their motor systems. Specifically, cognitive or neurological deficits affect the
integrity of information processing or reduce the efficiency of
information delivering in the sensorimotor systems. The increased noise may result in slower movements and a greater
variability in force generation.
Studies have shown that Tai Chi, a traditional form of Chinese
fitness exercise, improves cardiovascular fitness while reducing
the stress, anxiety, and depression levels of its participants (e.g.,
14,15). Tai Chi's slow, smooth, and graceful movements, and its
demands for mental concentration during exercise are assumed
to enhance physical and psychological well-being (16). In addition, compared to resistance training or locomotor activities, Tai
Chi practice results in better balance control (17,18) and
smoother aiming arm movements for senior adults (18). The
possible explanation for the effectiveness ofTai Chi practice in
movement control is that Tai Chi improves senior adults' sensorimotor integration. For instance, compared to participants in
other physical activities (e.g., walking or jogging), Tai Chi participants may better remember and prepare movements or they
may use feedback more effectively, leading to improved movement control (18).
Manual movements such as reaching, grasping, transporting,
and/or manipulating objects are essential to everyday life.
Consistent arm movements contribute to quality of life in general, as well to some specific daily activities. For example,
cooking, dining, dressing, and/or driving require normal functions of hand and arm movements. Typically, senior adults' arm
movement control diminishes, which influences their everyday
activities (3). Even though studies suggest that practice reduces
force variability in older adults (e.g., finger resistance training,
a laboratory-oriented motor task) (19), less is known about the
effects of real life physical activities on reducing force variability for seniors. From a practical standpoint, a legitimate
question is whether Tai Chi practice or participation in more
traditional locomotor activities can enhance senior adults' consistency of force generation in aiming arm tasks.
This study compares force outcome variability (fluctuation
of downward pressure) between Tai Chi and locomotor activity
groups in linear and curvilinear aiming arm movements.
Several justifications seem reasonable for this experimental approach. First, both locomotor activity and Tai Chi are nonresistant and total body exercises (arms, legs, and trunk). On one
hand, locomotor activities such as walking and jogging are the
most common forms of physical exercises for senior adults. On
the other hand, Tai Chi, a whole body exercise, is recommended
for seniors (16,18). Some previously published studies have
used locomotor activity as a control or comparison condition to
examine the influence ofTai Chi on seniors' physical and mental well-being and movement control (e.g., 15,20,21). Thus, it is
rational to compare the effects of these two physical activities
on force control. Second, a variety of manual movements are
M629
M630
YAN
performed in our daily activities. In this study, linear and curvilinear manual tasks were used to assess the effects of Tai Chi or
locomotor activity on the force control for arm/hand performance. Third, the arm movements used in the study are driven
by the resultant force of horizontal and vertical planes. Because
the downward pressure of arm movement on the contacting
surface yields information regarding force fluctuation and magnitude during the movement (22), vertical pressure was the
dependent measure to examine force control during arm movement execution.
After 8 weeks of Tai Chi practice, senior Tai Chi participants
were hypothesized to better improve their consistency of force
output during aiming arm movements than the participants who
engaged in a form of locomotor activity (walking or jogging).
Tai Chi's smooth, continuous, and well-coordinated arm and
whole body movements may facilitate senior adults' integrating
perceptual-motor information or reducing noise in their sensorimotor systems (e.g., effectively memorizing, planning, or executing movements), which may result in better control of force
production in aiming arm movements.
METHOD
Participants
Twenty residents from a nursing home, ages 76 to 88 (M =
78 years, SD = 2.3), volunteered to take part in the study
(Appendix, note 1). A survey conducted prior to the study indicated that all participants were independently ambulatory, righthanded, had normal or corrected to normal vision, and had no
severe musculoskeletal or/and neurophysiological disorders
(e.g., arthritis, hand tremor, or Parkinson disease). They also
had no prior experience with the manual tasks in the study. Tai
Chi participants had minimal exposure or experience with Tai
Chi at the beginning of the study. Most of the participants did
not regularly engage in physical activities (e.g., exercising with
a low frequency, short duration, and/or inconsistent schedule).
All participants passed the verbal part of the Mini-Mental
State Examination (23), suggesting that the cognitive or mental
capabilities of these individuals were normal (e.g., orientation,
memory, and attention). These residents chose and participated
in either one of the two exercise groups: Tai Chi (n =12, M =
79.3, SD = 2.4; three males and nine females; weight M = 70
kg, SD = 5.8; height M = 173 em, SD = 3.7) and locomotor activity (walking or jogging, n = 8, M = 79.5, SD = 1.9; two males
and six females; weight M = 68 kg, SD = 5.3; height M = 171
ern, SD = 3.8) (Appendix, note 2). Informed consent was
obtained from each participant prior to the research. All these
individuals completed at least 80% of the Tai Chi practice or
locomotor exercise sessions.
Apparatus and Procedures
A WACOM digitizer (UD-1218-R and a stylus pen: sampling
rate at 205 Hz, 0.25 mm spatial accuracy, maximum 250 g/cm?
vertical pressure) was used to measure arm movement performance. A portable personal computer (Sony PCG-717C) was
connected to the digitizer displayed and recorded the arm
movements (displacement, time, and downward pressure data).
The participants either engaged in Tai Chi practice or a lowintensity locomotor activity (walking or jogging) for 8 weeks: 3
times per week, 4.5 minutes per session, while under the super-
vision of trained personnel. Because the 24-form simplified Tai
Chi practice (a modified form of Yang's Tai Chi) is relatively
easier to learn and control exercise duration and intensity than
other styles (16,24), it was used in this study. According to the
American College of Sports Medicine (ACSM) guidelines for
exercise testing and prescription (25), a low-intensity exercise
refers to a physical activity that produces a less than 60% of
maximum heart rate (HR) for its participants. Thus, the HR was
controlled at an average level of 100 beats/min for all the activities involved. Manual movement tests were conducted before
the intervention (pretesting), at week 4 of the practice/exercise
(retest), and at the end of the practice/exercise (post-testing).
Each participant was tested individually and given standardized
instructions for the tests.
Each participant sat in front of the digitizer placed on a tabletop. With the right hand, each participant held a stylus pen in a
writing style and performed the linear and curvilinear tasks in a
counterbalanced order. After the "start" signal was given, the
participant moved a hand-held stylus from the starting position
to the target as rapidly and as accurately as possible, and with
minimum and consistent vertical downward pressure (Appendix, note 3). In the linear task, each participant moved the
pen directly to the target (a forward arm motion in the midline
of the body). In the curvilinear aiming task, each participant
moved the pen in a curved line, passing the stylus through a circle that was located midway between the starting position and
the target (8 em off the straight line), and then stopped inside
the circle of the target.
To prevent any body movements, each participant was required to keep the chest against the table during the arm movements. Only the successful trials (the hand-held stylus moved
into the circle of the target) were analyzed to standardize the
motion performed, and reduce the movement intertrial variability for each participant (Appendix, note 4). The distance between the target and the starting position was 20 em (about 60%
of the arm length), and the diameter of the target/home circle
was 2 cm. Each participant had a total of five to seven practice
trials (to fully understand the tasks), followed by 20 data trials
for each task (the means of data trials were used for the analyses). Figure 1 shows the linear and curvilinear manual tasks.
Data Analyses
The raw data were filtered using a Butterworth digital filter
with a cutoff frequency of 8 Hz. Movement position (x, y axes),
time, and pressure data (z axis, downward direction) were then
reduced for the analyses (see Appendix, note 1). The computer
software registered movement "onset" when the velocity exceeded 4 mm/s. The movement "offset" was marked as the arm
movement stopped inside the target (the velocity at the last sampling point declined below 4 mm/s). Movement duration or
time (MT) was the time from the "onset" to the "offset" of the
arm movement.
As soon as the stylus touched the digitizer, the arm movement
produced a continuous downward pressure on the surface of the
digitizer until the stylus departed from the digitizer. A standard
deviation (SD) of 20 data points of pressure between the "onset"
and "offset" of movement (movement duration or MT) was used
to measure the overall variability in arm movement force output.
For each trial, 20 data points of pressure were sampled in terms
of equal MT among data points (Appendix, note 5). Then for
TAl CHI REDUCES MOVEM£NT FORCE VARIABILITY
each trial a SD was computed for these 20 data points . Because
the MTs were slightly varied among the participants, in any
given trial the pressure variability was standardized: the SD of
20 data pointslMT. The average standardized SD for 20 data trials was used as a dependent measure for overall pressure variability. A smaller standardized SD represented less variability in
pressure during the execution of arm movement.
An analysis of variance (ANOYA) with repeated measures
was used to analyze the data. The independent variables were
the type of exercise (Tai Chi vs locomotor activity) and the task
(linear vs curvilinear task) . The repeated-measures factor was
the testing session (pretest, retest, and post-test) . The dependent
measure was overall variability in arm movement pressure. No
examination was conducted for the gender effect because no
gender-associated differences have been identified in senior
adults' arm movements (3).
REsULTS
Means and SDs for overall variability in movement pressure
(g/cm-) between Tai Chi and locomotor activity among the
tasks across three test sessions are displayed in Figure 2. There
were no significant three-way interactions (exercise type by
task by test session), two-way interactions, nor main effects of
exercise type and test session. However, the effect of task was
M631
=
significant, F( I,18) 34.43, p < .00 I . The participants had a
higher pressure variability in the curvilinear task than in the linear task. Furthermore, within the curvilinear arm movement
task, the interaction of Exercise Type by Test Session, F(2, 17)
= 7.79, P < .0 I; the main effects of exercise type, F(I , 18) =
6.32,P < .05; and test session, F(2, 17) = 25.62, p < .00 I, were
significant. The results suggest that in the curvilinear arm
movement task (the top panel, Figure 2), Tai Chi participants
demonstrated considerably more reduced variability in movement pressure than their counterparts in the locomotor activity
group. Because the linear task reflected a trend in training effect, the data were still reported even though no statistical significance was found (the bottom panel, Figure 2).
Figure 3 demonstrates four characteristic pressure profiles of
the curvilinear arm movements (Appendix, note 6). These samples indicate that after Tai Chi practice, the participant executed
the curvilinear arm movement with increased consistency in
controlling force output, whereas the participant in the locomotor activity did not substantially reduce the inconsistency. These
data support the prediction that Tai Chi practice is a better form
~
-
«l
35
.9
>- :Jl
Q Target
~
...III
>
...41
.....
::J
...Gl
a.
o
25
III
31
15
10
5
o
Groups
Home
Linear Ann Movement
O.~arget
'.
0
0 ····
Home
~
-
«l
35
.9
>- :Jl
~
EJ
~Ter;t
§
PeS
25
III
...III
>
...41
.....'"
31
15
::J
10
41
5
a.
o
Groups
Curvilinear Ann Movement
Figure 1.The diagram of linear and curvilinearann movements(top view).
Figure 2. The interactionof ExerciseType by TestingSession for the variable
of pressurevariability (top panel:curvilineartask; bottom panel:lineartask).
YAN
M632
of exercise for stabilizing force outcome during manual aiming
movements.
DISCUSSION
Research on aiming arm movement control is primarily interested in the processes that underlie force production as well
as in the factors that cause variability in force production (9,10).
Studies have reported that in addition to inconsistent force production in senior adults' motor performance, aging is associated
with slower, more variant, and less smooth motor responses
(3-5,7). Cognitive deficits or noise in senior adults' motor systems may contribute to, in part, their decline in motor performance (11-13). At issue in this study was whether Tai Chi practice could enhance its participants' sensorimotor integration or
reduce cognitive deficiency by better preparing and coordinating
the arm and total body movements.
The results of this study suggest that relative to locomotor
activities, Tai Chi is a better form of exercise for improving the
consistency in senior adults' force output during the curvilinear
arm performance (Figure 2, the top panel; Figure 3). One possible explanation for this finding would be that in comparison
with exercise of walking or jogging, Tai Chi involves more total
body exercise as well as more brisk arm movement training,
leading to a better practice effect. For instance, a Tai Chi routine includes a variety of continuous curvilinear arm movements, as well as the combination of linear and curvilinear arm
movements. This may explain why after 8 weeks of exercise,
Tai Chi participants considerably decreased force variability in
the curvilinear manual movement to a greater degree than the
senior individuals who were engaged in a locomotor activity
(Figures 2 and 3). These data demonstrate the benefits of an alternative exercise (e.g., Tai Chi) for reducing seniors' variability
in force output that were not previously reported in the studies
of aging and physical activity.
Tai ChilPost-test
Tai Chi/Pre-test
ZD
1.0
u
Time (sec)
Time (sec)
LocomotorlPost-test
Locomotor/Pre-test
N'
E
-•..
(,)
,g
:J
en
to
•
e,
~
1.&
Time(sec)
Q&
'.0
Ti me(sec)
Figure 3. Representative profiles of pressure variability for the participants in each treatment group and under pre- and post-testing conditions (only the
curvilinear task).
TAI CHI REDUCES MOVEMENT FORCE VARIABILITY
M633
Furthermore, from a perceptual motor control perspective,
several characteristics of Tai Chi may explain its beneficial effects on force control during manual aiming movements. For
example, Tai Chi's fundamental characteristics such as its demand for mental concentration over its slow, smooth, selfpaced, and coordinated whole body movements, along with its
requirement for remembering and understanding body orientations, movement patterns, and the sequence of the movement
routine (16), provide the Tai Chi participants a unique exercise
in utilizing sensory information for movement control.
Particularly, Tai Chi practice focuses on learning and practicing the body motions in a progressive sequence. A Tai Chi performer has to understand and remember all movement components in order to complete the whole routine. In addition, the
movements are performed slowly, smoothly, and in a selfcontrolled rhythm. These movement characteristics direct a performer's attention to the sensory information that the performer
can integrate into the controlled arm and total body movements.
Because the Tai Chi movements are slow and the performers
focus on the performance, the participants, especially the senior
adults, would have adequate time to process sensorimotor information for body movements. Alternatively, as a function of
Tai Chi practice, the senior participants might reduce cognitive
deficits or noise by better response preparation that was reflected in improved force control in the manual movement
tasks. Thus, strengthening senior adults' use of sensory feedback may be an essential factor for force control.
However, all participants had higher force variability in the
curvilinear task than in the linear task (Figure 2). The curvilinear task was more complex than the linear task (controlling
movement path or changing movement direction during the execution)' leading to a higher force variability (26). The results
suggest that manual movements involving higher task demands
or constraints may decrease the performance level of seniors.
To maintain relatively stable motor functions or performance,
reducing the degree of task complexity can be a useful strategy
for older adults.
In short, the findings from this study suggest that Tai Chi
practice may serve as an alternative physical activity for improving senior adults' force control in aiming arm movements.
Tai Chi's deep concentration could facilitate perceptual motor
integration (movement preparation and execution), and lead to
consistent force production for the curvilinear arm movements.
Practically and theoretically, extensive Tai Chi practice may improve the quality of life of senior adults as well as enhance their
manual movement functions.
4. Spirduso WW. Reaction time and movement time as a function age and
activity level. J Gerontal. 1975;30:435-440.
5. Welford AT. Between bodily changes and and performance: some possible reasons for slowing with age. Exp Aging Res. 1984;2:73-88.
6. Cole KJ. Grip force control in older adults. J Motor Behav. 1991;
23:251-258.
7. Cole KJ, Beck CL. The stability of precision grip force in older adults.
J Motor Behav. 1994;26:171-177.
8. Darling WG, Cooke ro. Changes in the variability of movement trajectories with practice. J Motor Behav. 1987;19:291-309.
9. Schmidt RA, Zelaznik HN, Hawkins B, Frank JS, Quinn JT. Motor output
variability: a theory for the accuracy of rapid motor acts. Psychol Rev.
1979;86:415-451.
10. Newell KM, Corcos DM. Issues in variability and motor control. In:
Newell KM, Corcos DM, eds. Variability and Motor Control. Champaign,
IL: Human Kinetics; 1993:1-12.
II. Johansson RS. How is grasping modified by somatosensory input? In:
Humphrey DR, Freund HJ, eds. Motor Control: Concepts and Issues.
New York: Wiley; 1991:78-88.
12. Kinoshita H, Francis PR A comparison of prehension force control in
young and elderly individuals. Eur J Appl Physiol. 1996;74:450-460.
13. Walker N, Philbin DA, Fisk AD. Age-related differences in movement
control: adjusting submovement structure to optimize performance.
J Gerontal: Psychol Sci. I997;52B:P40-P52.
14. Brown DD, Mucci WG, Hetzler RK, Knowlton RG. Cardiovascular and
ventilatory responses during formalized T'ai Chi Chuan exercise. Res Q
Exerc Sport. 1989;60:246-250.
15. Jin P. Efficacy ofTai Chi, brisk walking, meditation, and reading in reducing mental and emotional stress. J Psychosomat Res. 1992;36:361-370.
16. Yan JR. The fitness and health benefits ofTai Chi. J Phys Edu Rec Dance.
1995;NovemberlDecember:61--63.
17. Wolfson L, Whipple R, Derby C, et al. Balance and strength training in
older adults: intervention gains and Tai Chi maintenance. J Am Geriatr
Soc. 1996;44:498-506.
18. Yan JR. Tai Chi practice improves senior citizens' balance and arm movement control. J Aging Phys Activ. 1998;6:271-284.
19. Keen DA, Yue GH, Enoka RM. Training-related enhancement in the
control of motor output in elderly humans. J Appl Physiol. 1994;77:
2648-2658.
20. Judge JO, Lindsey C, Underwood M, Winsemlus D. Balance improvements in older women: effects of exercise training. Phys Ther. 1993;
73:254-262.
21. Schaller KJ. Tai Chi Chih: an exercise option for older adults. J Gerontal
Nurs. 1996;22:12-17.
22. Wann J, Nimmo-Smith I. The control of pen pressure in handwriting: a
subtle point. Human Movement Sci. 1991; 10:223-246.
23. Folstein MF, Folstein SE, McHugh PR Mini-mental state: a practical
method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12:189-198.
24. Yan JH, Downing JR. Tai Chi: an alternative exercise form for seniors.
J Aging Phys Activ. 1998;6:363-375.
25. American College of Sports Medicine. ACSM's Guidelines for Exercise
Testing and Prescription. 5th ed. Philadelphia: Williams & Wilkins; 1995.
26. Christina RW. The 1991 C. H. McCloy research lecture: unraveling the
mystery of the response complexity effect in skilled movement. Res Q
Exerc Sport. 1992;63:218-230.
ACKNOWLEDGMENTS
Received July 21, 1998
Accepted April 7, 1999
The kinematics of arm movement are pnblished in the Journal ofAging and
Physical Activity, 1998,6,271-284.
Address correspondence to Jin H. Yan, PhD, Department of Health and
Human Performance, The University of Houston, Houston, TX 77204-6321.
E-mail: [email protected]
REFERENCES
1. Darling WG, Cooke JD, Brown SH. Control of simple arm movements in
elderly humans. Neurobiol Aging. 1989;10: 149-157.
2. Goggin NL, Meeuwsen HJ. Age-related differences in the control of spatial aiming movements. Res Q Exerc Sport. 1992;63:366-372.
3. Yan JH, Thomas JR, Stelmach GE. Aging and rapid aiming arm control.
Exp Aging Res. 1998;24:55--69.
Appendix
1. Initially 68 seniors signed up for this research project. Data were
excluded for the residents who did not pass the Mini-Mental State
Examination (14 individuals, 20%) and who missed 20% of the Tai Chi
practice or locomotor activity sessions (additional 16 individuals, 24%).
Kinematic data of curvilinear manual task and the results of balance
testing for these 38 participants were reported in Yan (18). In addition,
participants who had missing data points in vertical pressure of either
linear or curvilinear manual tasks were not included in the next part of
research (additional 18 individuals, 26%). Therefore, only 20 seniors
M634
remained in the present study, which examined the effects of Tai Chi on
kinetic characteristics of linear and curvilinear manual tasks.
2. Considering the age of the participants (from 76 to 88 years of
age), their exercise forms (80% of the residents irregularly participated
in walking or jogging), and their preference of activity intervention, to
facilitate residents' active engagement in the study and maintain a relatively low drop-out rate, each resident was allowed to select hislher
own activity in this study (either Tai Chi or a locomotor activity).
3. The interval between the "start" signal and the "onset" of the
arm movement (RT) ranged from 280 to 350 ms. After five to seven
practice trials, most seniors understood the task requirements. There
were not significant differences in the error rates between groups: Tai
Chi, linear task, M = 18%, curvilinear task, M =23%, Locomotor activity, linear task, M =]6%, curvilinear task, M =22%.
4. Two major aspects of difference were observed between the unsuccessful (missing the target) and successful (hitting the target) trials
YAN
in the manual movements. Besides missing the target, typically the unsuccessful trials were inconsistent and awkward, resulting in a greater
variability in the movement paths than the successful trials. To compare similar movement characteristics between Tai Chi and locomotor
activity, only the successful trials were analyzed.
5. MT ranges: from 500 to 800 ms for the linear task and from
]200 to 1800 ms for the curvilinear task. Certainly more data points of
pressure could better reflect the differences in pressure variability.
Because the interval among data points was very short: 60 to 100 ms
by using 20 data points, the use of the SD of 20 data points could be
considered reasonably accurate in capturing the variability of pressure
during arm movement execution.
6. The trials that had the closest values to the group means of pressure and MT were selected as the typical trials.