| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Reports |
RC Bakken, PT, MSPT, is Physical Therapist, Fairview Southdale Hospital, Edina, Minn. Address all correspondence to Ms Bakken at 2649 Toledo Ave S, St Louis Park, MN 55416 (USA) (curbak{at}yahoo.com)
JR Carey, PT, PhD, is Associate Professor and Director, Program in Physical Therapy, University of Minnesota
RP Di Fabio, PT, PhD, is Professor, Program in Physical Therapy, University of Minnesota
TJ Erlandson, PT, MSPT, is Physical Therapist, Orthopedic Rehabilitation Specialists Inc, Minneapolis and St Louis Park, Minn
JL Hake, PT, MSPT, is Physical Therapist, Courage Center, Golden Valley, Minn
TW Intihar, PT, MSPT, is Physical Therapist, Innovations Rehabilitation, Sparta, Wis
Submitted January 10, 2000;
Accepted March 27, 2001
 | Abstract |
|---|
Key Words: Aerobic exercise • Elderly • Information processing • Tracking
| Introduction |
|---|
 TopAbstract Introduction MethodsResultsDiscussionConclusionReferences |
|---|
Reaction-time paradigms have long been used to study information processing, which refers to the cognitive activity that leads to a purposeful movement in response to a perceived stimulus.13 Spirduso14 considered fast reaction time to be an indicator of good health and showed that reaction times slow with aging, indicating that information processing declines with age. However, the speed of information processing differs within age groups, and it has been postulated15 that factors such as physical fitness may explain this variability. Researchers15,16 have found that people who exercise regularly as part of their lifestyle tend to have faster reaction times than those who are sedentary. However, others17,18 argue that those who have chosen to exercise regularly may have done so because they were genetically endowed with better-than-average motor skills and faster reaction times.
Reaction times appear to be faster in sedentary subjects (those who reported no regular exercise on a questionnaire) who participated in an exercise program than in those who did not. In a study of 80 older subjects (aged 50–75 years), Lord and Castell11 demonstrated improved reaction times in subjects who received 10 weeks of aerobic activities 2 times per week compared with controls. Clarkson19 similarly showed improved reaction times in elderly active subjects who received 16 weeks of aerobic training compared with inactive controls. Dustman et al20 found improved reaction times in subjects aged 55 to 50 years following 4 months of aerobic exercises compared with controls. In a study of 48 women aged 57 to 85 years, Rikli and Edwards21 found improved reaction times in subjects who performed aerobic exercises 3 times per week for 3 years, whereas reaction times of control subjects worsened. Identical twins with different exercise histories showed improved reaction times for exercisers when compared with their genetically identical but nonexercising twin.17 Not all researchers have shown that psychomotor performance improves in young adults and elderly adults with aerobic training.18,22 Researchers17,18 have postulated that uncontrolled variables such as disease, smoking, alcohol abuse, intelligence, education level, and genetic differences may account for these inconsistent results.
Although reaction time is frequently used to study information processing,23 it is a discrete motor skill.13 Functional activities such as combing hair, writing, and feeding require the ability to perform continuous motor skills.13 Tracking tasks have also been used to study information processing during controlled continuous movements.24 In adults, the ability to track declines with age.25–28 Numerous researchers29–32 have shown that reaction time corresponds to changes in information processing.
According to Schmidt,13 information processing involves the important cognitive steps of stimulus identification, response selection, and motor programming necessary to produce skillful movements. One method used to study information processing involves stimulus response (S-R) compatibility. Stimulus response compatibility is the degree of congruence between a stimulus and a required response.33 Noncomparable conditions, such as moving the right hand to a target in response to a stimulus in the left visual field, are unnatural and required a greater depth of information processing to execute the correct response. Conditions that challenge the mental transformation associated with a right-sided response to a left-sided stimulus result in less efficient performance of a perceptual motor task.34,35 Both tracking studies36,37 and reaction-time studies33,38 have shown impaired performance in elderly people under conditions of S-R noncompatibility. Thus, we used an S-R noncompatible position in this study because we believed that performance in the noncompatible position, with its inherent greater mental challenge, would be a more sensitive indicator of improved information processing.
Aerobic training has been shown to have a positive effect on discrete psychomotor skills in elderly people,12,15,21 but the effect of aerobic training on continuous psychomotor skills is unknown. Continuous skills comprise much of our daily activity, and if the information processing associated with these activities can be improved with aerobic training, activities of daily living also may improve.14,20,39,40 Therefore, the purpose of our study was to investigate the influence of an 8-week aerobic training program on continuous psychomotor skills, as measured by a finger-movement tracking test, in a group of elderly subjects. Our hypothesis was that an exercise group would show both finger-movement tracking and aerobic training effects when compared with a nonexercising control group.
| Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Prior to the initiation of the study, the primary physician for each subject provided medical clearance for participation in the study. All subjects gave informed consent at the beginning of the study.
Tracking Instrumentation and Procedure
For the tracking test, a Dell computer* displayed 4 cycles of a fixed and predictable sine wave at 0.4 Hz that served as a target for the subjects to track with the cursor.37 An electrogoniometer, housing a potentiometer,
was attached to the index finger metacarpophalangeal (MP) joint of the subject's dominant hand. An analog-to-digital converter
was used to sample the voltage signal from the potentiometer at 60 Hz. During a given tracking trial, the cursor moved horizontally across the screen with a 10-second sweep time; therefore, each trial was 10 seconds long. The subject was required to adjust the vertical position of the cursor to track the fixed-target sine wave using finger extension and flexion movements of the index finger MP joint. The cursor began its sweep 0.3 second before entering the actual target zone in an attempt to eliminate the confounding effects of reaction time.
Each subject sat in a hard-backed chair facing the computer screen with his or her eyes approximately 75 cm from and level to the computer screen. The subject's dominant hand rested on a forearm support, with the forearm midway between pronation and supination (Fig. 1). We used the dominant hand because it is the hand that is used most frequently for skillful activities, and we determined the dominant hand by subject report. The elbow was positioned at an angle of approximately 70 to 80 degrees of flexion, while the wrist was in a neutral position between flexion and extension. The proximal and distal interphalangeal joints of the index finger and the thumb were not restrained.
|
After maximal flexion and extension movements were recorded, the examiner explained the task. The examiner then did a demonstration. The examiner started the test and passively moved the subject's finger in the correct fashion to track the sine wave. Next, the subject performed 3 practice trials. During these 3 trials, the examiner offered advice and answered any further questions. The subjects were not given their scores, but they did see the line created by their response superimposed over the fixed sine wave. The same examiner recorded all trials for the pretest and the posttest. The subjects then performed three 10-second trials (tracking movements) with 30 seconds of rest between trials. No feedback or advice was given by the examiner. Data from these 3 test trials were recorded. This entire procedure was followed prior to the 8-week exercise program and then again within 5 days of completion of the exercise program (except for one subject—see "Data Analysis" section).
To score the tracking performance, the computer was programmed to calculate the root-mean-square (RMS) error between the target and the response.37 However, because the value of the target amplitude varied across subjects according to their own finger range of motion, this raw RMS error score could not be used to compare across subjects. Instead, the RMS error was normalized to each subject's own range of motion and converted to an accuracy index (AI).37 The maximum possible AI score is 100%. Negative values result when the tracking response occurs on the wrong side of the midrange line separating the extension and flexion phases of the sine wave.
Graded Exercise Tolerance Test Equipment and Procedure
The submaximal graded exercise tolerance (GXT) test is a step test designed for testing elderly or sedentary individuals to determine whether an aerobic training effect can be achieved with exercise.1 The GXT test followed the tracking test immediately in the pretest and posttest.
For the GXT tests, a set of adjustable parallel bars and 3 wooden steps were used. The steps were 45.72 cm long, 45.72 cm wide, and 10.16, 20.32, or 30.48 cm high. This test consisted of stepping in a specific pattern onto and off of a step at a frequency of 20 mounts per minute for 3 minutes at each stage of the test. The subjects were able to hold on to the parallel bars for balance while stepping, but could not lean on them. In stage 1 of the stepping pattern, both of the subjects' feet were on the ground, and the subjects simply stepped forward and backward on the level floor. Each subject took one step to each beat of a metronome that was set to count 80 times per minute (ie, right foot forward, left foot forward, right foot backward, left foot backward). After 3 minutes, the subject took a 1-minute break. In stage 2, the subject raised the right foot onto the 10.16-cm step, followed by the left foot. The subject proceeded to step down with the right foot, and finally down with the left foot. Stage 2 also lasted 3 minutes, followed by a 1-minute rest. In stage 3, the step was 20.32 cm high, and stage 4 involved a 30.48-cm step; both stages lasted 3 minutes with a 1-minute rest in between.
Before the test began, each subject's resting heart rate and blood pressure measurements were taken with the subject in a standing position. The heart rate was monitored with a Polar heart rate monitor (model 1902101)
with a chest strap. The heart rate was shown on a watch worn by the subject. All blood pressure measurements were taken by the same examiner with the same blood pressure cuff and stethoscope. During the rest period between stages, blood pressure and heart rate were recorded while the subject remained standing. Then, if the subject was able, the next stage began immediately with the next higher step. The GXT test was ended by the subject's decision that he or she had reached his or her maximal tolerance. The examiner ended the test, however, if the subject reached a heart rate of greater than 75% of his or her target heart rate or if the subject's systolic blood pressure rose above 225 mm Hg. For example, if a subject was 80 years age, the target heart rate range (THRR) was 108 to 127 mm Hg, and 75% of that THRR was 81 to 95 mm Hg. Therefore, if the heart rate rose above 81 mm Hg, the test was ended. The test was also ended if the subject reported or demonstrated exercise intolerance as indicated by shortness of breath; chest, neck, jaw, or arm pain; dizziness, nausea, or confusion; or sudden pallor or inappropriate sweating.1
To determine whether any aerobic training effect had taken place, rate-pressure product (RPP), defined as systolic blood pressure multiplied by heart rate, was used. A decrease in RPP is indicative of improved efficiency of cardiovascular mechanisms and is considered to be a quantitative measure of aerobic training.1 The finger-movement tracking test and the GXT test were used twice for both the control and experimental groups, once as a pretest before the exercise program and once as a posttest within 5 days after the end of the exercise program. The investigators who performed the testing were unaware of each subject's group assignment.
Intervention
The intervention consisted of group exercise conducted by 2 physical therapist students under the direction of a licensed physical therapist. An on-site emergency medical technician was available. The exercise sessions were held 3 times a week for 8 consecutive weeks for 1 hour each session. Attendance was taken at each session to monitor adherence.
Each subject's THRR was calculated as shown in Table 1,41 and pulse rates were continuously monitored using the heart rate monitors. Caution was taken to ensure that the subjects' heart rates did not exceed the upper limits of their THRR. In addition, each subject's blood pressure was periodically measured throughout the 8-week exercise program to ensure that the systolic value did not exceed 225 mm Hg.1 The exercise classes each consisted of a 10-minute warm-up period, an aerobic conditioning period that increased in duration systematically each week and also increased slightly in intensity, and a 10-minute cool-down period (Tab. 2).
|
|
The aerobic conditioning period consisted of calisthenics, stationary bicycling, and walking. The calisthenics consisted of repetitions of gross movements of the upper and lower extremities through full range of motion. The exercises included elbow flexion, shoulder flexion, marching in place, and knee bends, all in a standing position. Walking initially occurred on a carpeted level surface and progressed to a "walking course," consisting of approximately 30 m of a carpeted level surface with a decline ramp and one flight of stairs. A stationary bicycle was also used during the aerobic conditioning period. Each week, the aerobic conditioning period increased in duration and intensity in response to the subjects' increasing ability to exercise. A summary of the exercises and weekly progression is shown in Table 2. During the exercise sessions, subjects were instructed to add rest periods, to decrease repetitions, or to decrease the walking pace if they exceeded their THRR or if any abnormal pain or fatigue was experienced.
During the 8-week exercise period, the control group continued their normal, everyday routine, which did not include any aerobic exercise according to subject report. The exercise group was also instructed to avoid any additional exercise outside of the regular sessions.
Data Analysis
The data from 10 of the 15 subjects who entered the study were used for data analysis. In the experimental group, 2 subjects were eliminated due to illness and 1 subject was eliminated due to a lack of attendance (less than 50% of the sessions) in the exercise program. One additional subject in the exercise group became ill with influenza right after the completion of the exercise program, making it impossible to test him at the same time as the others. We were able to test his finger-movement tracking performance when he was feeling somewhat better, 2 weeks after the completion of the exercise program. However, by the time he fully recovered from his illness, so much time had passed since the completion of the exercise program that we believed a GXT posttest at this time would be invalid. Therefore, we included his finger-movement tracking pretest and posttest, but no GXT tests.
Two of the 7 control subjects were eliminated to allow for an equal gender balance and handedness in each group. We then created matched pairs of subjects in each group based on rank order of mean pretest tracking scores to enable comparison of subjects with equivalent tracking skills. The change in AI score from pretest to posttest was analyzed with a matched-pairs t test.
A commercial statistical software package|| was used for all data analyses. A chi-square analysis was done to test for gender proportions between groups and a 2-sample t test was done to check for a difference in age between the control and experimental groups. The dependent variable of primary interest was the mean AI score for the 3 tracking trials in the pretest and posttest. Reliability for the AI was determined by comparing the mean AI score from pretest to posttest in the control group using the intraclass correlation coefficient (ICC) (3,k).42 The ICC equation used in this study was: ICC (3,k) = (BMS – EMS)/BMS, where BMS is the before-target mean square and EMS is the error mean square. The ICC (3,k) for the average AI score was .89. To examine for training effects, one-tailed matched-pair t tests were performed comparing the change from pretest to posttest between the 2 groups in AI, RPP at rest, and RPP during the GXT.
Similar to Amundsen et al,1 we used work rates in comparing submaximal GXT responses from pretest to posttest. That is, if the highest exercise stage that a subject could complete (3 minutes) was stage 2 on the pretest, then the responses at the end of stage 2 on the posttest were those used for comparison even if the subject progressed to stage 3. Statistical significance was set at P<.05.
| Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
Figure 2 shows the mean AI scores for the 2 groups during the pretest and posttest. For the experimental group, the mean AI score increased from –16.0% (SD=33.5%) on the pretest to 10.1% (SD=23.4%) on the posttest, whereas for the control group, the mean AI score decreased from –2.7% (SD=27.5%) on the pretest to –16.2% (SD=37.3%) on the posttest. This change in mean AI scores from pretest to posttest was different between the 2 groups (t[4]=2.81, P=.02). Figure 3 shows tracking responses for an experimental subject, demonstrating the improved performance from pretest to posttest.
|
|
|
| Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
The average AI scores for both groups were considerably below those of a previous study that included elderly subjects.25 This finding could be due to the fact that the average age of our subjects was higher than the average age of the subjects in that study, and it is known that information processing declines with advancing years.39 In addition, the subjects in the previous study all lived independently in the community, whereas the subjects in our study all resided in a senior housing complex. Our results must be viewed with caution due to the very small number of subjects in what must be considered a pilot study.
Our subjects did not show an aerobic training effect in the experimental group. One possible reason for the absence of an aerobic training effect is that the exercise period may have been too short. The exercise period for our study was only 8 weeks long, whereas most other researchers have found positive cardiovascular changes with longer exercise programs, such as 16 weeks20 and 26 weeks.5 However, we chose an 8-week duration because Amundsen et al1 showed that it was possible to obtain an aerobic training effect in elderly subjects with as little as 8 weeks of exercise.
Another possible reason for the absence of an aerobic training effect was the low intensity of the exercise program. Because of our subjects' age, their THRRs were low. Subjects were often instructed to reduce their intensity of exercise because their heart rate exceeded the recommended target heart rate. This low-intensity exercise was chosen for the subjects' safety because of their advanced age.1
Because medication use did not preclude participation in this study, the massed use of various medications by all subjects may also have contributed to the lack of an aerobic training effect. Many medications taken by the subjects have a limiting effect on heart rate and blood pressure, which, in turn, could have affected the RPP, the primary measure we used to determine whether a training effect occurred.
The number of subjects in our study was small, thus limiting statistical power. The number of subjects was determined by a convenience sample. Recruitment of a larger number of elderly subjects who were without medical problems and who would commit to a lengthy duration of regular exercise was difficult. Why this would obscure determination of an aerobic effect but not a change in tracking, however, is not clear.
Kramer et al43 studied the effects of exercise on function in 124 elderly human subjects. These subjects underwent either aerobic exercise (walking) or anaerobic exercise (stretching) over a period of 6 months. For the subjects who received aerobic training, the results revealed selective improvements in reaction time. Of particular interest, the authors also stated that the improvements they found required only a small increase in aerobic fitness. Thus, although speculative, this finding of Kramer et al,43 in combination with our finding of improved tracking without an aerobic training effect, suggests that the improvement in information processing may be coupled not so much to "aerobic" training but perhaps to some other event related to the training, such as repetitive neuronal firing. Neurotrophins have been reported to be important for the function and survival of many neurons in the brain,44–46 and investigators have found that wheel running in laboratory animals increases neurotrophin levels in the brain.44–47 Indeed, Gomez-Pinilla et al47(pp6-7) contended that increased levels of neurotrophins induced by neural activity provide a molecular basis for maintaining neuronal function and may be useful for the development of strategies to prevent decline of function following pathology or aging and to promote rehabilitation following brain trauma.
Mechanisms for Improved Information Processing
Whether exercise can sustain or improve the performance of skills seemingly unrelated to the exercise is very relevant to physical therapy. As just one example that relates to both orthopedic and neurologic areas of rehabilitation, in elderly people with a newly fractured hip, it is inviting to consider whether repetitive upper-extremity exercise (eg, cycling) during the convalescent period could forestall the cognitive and motor decline that frequently accompanies this clinical problem.
| Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
| Footnotes |
|---|
This study was approved by the University of Minnesota's Committee on the Use of Human Subjects.
This project was completed in partial fulfillment of the requirements for the Master of Science in Physical Therapy degree for Ms Bakken, Mr Erlandson, Ms Hake, and Mr Intihar.
* Dell Computer Corp, One Dell Way, Round Rock, TX 78682. 
Waters Manufacturing Inc, Longfellow Center, Wayland, MA 01778.
Interactive Structures Inc, 146 Montgomery Ave, Bala Cynwyd, PA 19004.
Polar Electro Inc, 370 Crossways Park Dr, Woodbury, NY 11797-2050.
|| NCSS Statistical Software, 329 North 1000 East, Kaysville, UT 84037.
| References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
