| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a Balance Research Laboratory, Department of Kinesiology, University of Lethbridge, Alberta, Canada
b Gait and Posture Laboratory, University of Waterloo, Ontario, Canada
Lesley A. Brown, Department of Kinesiology, University of Lethbridge, Lethbridge, AB, Canada T1K 3M4 E-mail: l.brown{at}uleth.ca.
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Methods. Fifteen healthy older adults (mean age = 69.53 ± 5.78) and 15 younger adults (mean age = 22.00 ± 2.17) performed Brooks' Spatial Letter Task (BST) while standing in four conditions of postural threat. Galvanic skin conductance (GSC) was collected to measure changes in physiological arousal. BST task duration was used to measure changes in cognitive performance, and center of pressure (COP) area was used to determine changes in postural stability across each of the testing conditions. A prioritization index was calculated based on the relative change in BST and COP across testing conditions. This measure was used to quantify changes in the relationship between postural control and secondary task performance under conditions of postural threat.
Results. Measures of GSC increased in younger and older adults in response to conditions of increased postural threat. Prioritization index scores revealed that a greater number of older adults than younger adults prioritized postural control over secondary task performance under conditions of increased postural threat.
Conclusions. Environmental contexts that alter postural threat may lead to modifications in the prioritization between postural control and secondary task performance among older adults. This relationship may have implications for postural control under situations that do not afford deterioration in secondary task performance.
IT is now well established that postural control requires cognitive resources (1). It is also known that greater cognitive resources are necessary as a postural task becomes more difficult, such as when walking or balancing on one foot (2). Finally, and most compelling, research evidence suggests that older adults (OA) require greater cognitive resources than young adults (YA) when performing the same postural tasks (3)(4). Thus, because postural control relies on cognitive resources, performing concurrent cognitive tasks may be particularly deleterious among OA (5) or among pathological populations who may be challenged by cognitive deficits (6).
Recent research findings suggest that there is a dynamic interplay between cognition and postural control and that secondary task performance can be prioritized over postural control in conditions that do not threaten balance (7)(8). For example, Maylor and Wing (8) examined the effects of different types of cognitive tasks on postural control and demonstrated that tasks that interfered with visual-spatial components of cognitive processes were most detrimental to postural stability. In addition, it was also revealed that the introduction of a secondary cognitive task had greater effects on balance among OA than among YA. In a similar study, Shumway-Cook and colleagues (7) examined postural control during concurrent performance of a sentence completion task and a judgment of line orientation task in YA and among OA with and without a history of falls. Contrary to predictions that postural control would be maintained at the expense of secondary task performance, it was revealed that regardless of fall history, postural stability was significantly reduced when performing the sentence completion task. Interestingly, however, balance was most impaired by the secondary task among OA who had a history of falls. These results imply that individuals with impaired postural control (e.g., OA with a history of falls) are more likely to lose their balance when performing a concurrent secondary task than individuals with normal balance. Indeed, a study by Lundin-Olsson and colleagues (9) determined that many OA in nursing homes were unable to walk while maintaining a conversation.
Although the results of previous studies indicate that postural stability may be neglected to maintain secondary task performance, it is possible that the prioritization of postural control may depend on environmental factors that alter the level of postural threat, such as walking on an icy sidewalk or descending a flight of stairs (7). To explore whether postural control is altered by postural threat, several studies have monitored postural control under different environmental contexts. Research to date has confirmed that postural control is modulated according to the degree of postural threat with improvements in postural control occurring in the most threatening environmental contexts (10)(11)(12)(13). However, although individuals appear to prioritize postural control under conditions of postural threat, it remains to be determined whether this apparent prioritization toward postural control is preserved during secondary cognitive task performance. Thus, the purpose of this study was to examine the relationship between secondary cognitive task performance and postural control under environmental conditions that altered postural threat. Furthermore, because differences exist in cognitive capacity (14) and fall risk (15) between YA and OA, a secondary purpose of this study was to investigate whether age-related differences existed in the propensity for a prioritization of postural control under conditions of postural threat.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Before testing commenced, all participants voluntarily provided informed consent according to guidelines of the Human Research Ethics Committee of the University of Lethbridge. In addition, participants were asked to complete a Falls History form that assessed fear of falling (scale 1 [not afraid] to 10 [very afraid]), fear of heights (y or n), and time since last fall (months). Participants also completed both the Gait Efficacy Scale (GES) (16) and the Activities-Specific Balance Confidence (ABC) Scale (17). These questionnaires were designed to assess participants' perceptions of their balance and their ability to perform daily activities. During testing, participants wore a tee-shirt or blouse, shorts, socked feet, and a safety harness over their clothes.
Manipulation of Postural Threat
An industrial hydraulic lift table (1.2 x 1.8 m; Pentalift, Guelph, ON) was used to alter the environmental context of the testing conditions. Participants were tested under two vertical height positions: Low (0.17 m) and high (1.4 m) from ground level and two position conditions on the lift table (middle and edge). Thus, four conditions of postural threat were included in this study: (i) low-mid (LM)—middle of the platform at ground level, (ii) low-edge (LE)—edge of the platform at ground level, (iii) high-mid (HM)—middle of the platform and elevated, and (iv) high-edge (HE)—edge of the platform and elevated. Middle and edge conditions were created to produce conditions of threat that did or did not permit an individual to step forward to recover his or her balance. The four different conditions modified the level of postural threat imposed upon the participants. Condition One (LM) was least threatening, and Condition Four (HE) provided the most postural threat. Fig. 1 provides illustrations of the four testing conditions used in this study.
|
Procedure
Participants were seated in chairs on the ground to receive instructions regarding the testing procedure and protocol for Brooks' Spatial Letter Task (BST) (19). Participants were presented with 16 different 20.32 cm x 27.94 cm cards showing one of eight large simple block letters (G, H, J, M, S, W, Y, or E) approximately 25.4 cm tall. Two cards were made for each letter so that each card had an asterisk printed on either the bottom-left or top-right corner of the letter to indicate the "starting point" for the initiation of the BST task (see Fig. 2). Participants were allowed to view the letter until they felt they had a clear mental image of it. Upon participant consent, the letter was taken away, and participants were asked to start classifying the letter, according to the set criteria, out loud from one of two predetermined starting points, and moving in a clockwise direction. Each corner of the letter (indicated by gray dots, Fig. 2) was to be classified with either a "yes" or a "no" according to set criteria. The criteria were either "top/bottom," indicating that any corners on the extreme top or bottom of the letter received a "yes" answer, or "left/right" indicating that the corners on the extreme left or right edges of the letter received a "yes" answer (see Fig. 2). Classification criteria and the presentation of letters were randomized so participants were unable to predict or memorize the answers for the task. A minimum of six practice trials were given to familiarize participants with the task, but testing did not proceed until the participants were comfortable and proficient at performing the task. Participant comfort was indicated by participant self-report, and task proficiency was dictated by a 90% minimum accuracy level on two consecutive practice trials.
|
Instrumentation
Forceplates were used to obtain ground reaction force and moment of force data necessary to calculate center of pressure (COP) in each condition. A headset with a microphone was worn by the participant and was used to collect audio data for the BST and PRT tasks. Audio data from the headset microphone were collected on a separate collection computer through an AW35 Pro Audio soundcard (sampling frequency = 22 KHz). Finger cuffs with silver/silver-chloride electrodes from a BioDerm Skin conductance Level Meter (UFI, Morro Bay, CA) were attached to the middle phalanges of digits three and four to collect Galvanic Skin Conductance (GSC) data.
Analog data were collected for 7 seconds during the PRT trials and for 15 seconds during the QS trials. Collection times for audio data in the BST trials varied depending upon the time required for the participant's performance of the task (mean = 14.19 ± 2.21).
Measures of Interest
Results from the GES and ABC questionnaires were compiled to assess perceptions regarding the ability to perform daily activities in the home and within the community for YA and OA. In addition, results from the Falls History questionnaire were analyzed to determine if OA and YA significantly differed on perceptions of fear of falling, fear of heights, and time since last fall. GSC data for each trial were expressed as mean values for comparative analyses. Verbal responses from the BST task were monitored, and error rates and BST performance duration times were recorded.
Custom-written algorithms were used to process all analog data (Matlab, The MathWorks, Natick, MA). Forceplate data were filtered using a four-order zero-lag Butterworth low-pass digital filter at a cut-off frequency of 5 Hz. Coordinates for the anterior/posterior (x) and medial/lateral (y) positions of the COP, relative to the forceplate origin, were calculated for the assessment of postural sway as follows:
COPx = My/Fz
COPy = Mx/Fz, where M = moment of force and F = force.
To normalize for differences in foot length and stance width, COP measures were expressed as a percentage of measured base-of-support for each subject (i.e., foot length for COPx and stance width for COPy). COP area was calculated for each condition as follows:
COParea(%BOS2) = COPx range(%BOS)* COPy range(%BOS).
Data were cropped according to the length of the shortest BST response duration to ensure that the duration of sway analysis was constant between subjects and to ensure that the analysis of COP area occurred during periods of verbal response for all subjects.
Data Analysis
Separate independent t tests were used to determine differences in the mean total score of the combined GES and ABC scores, fear of falling, and time since last fall scores between YA and OA. A chi-square test was used to determine statistical differences in the number of YA and OA who reported a fear of heights.
Due to technical limitations, data from five out of 15 OA and 11 out of 15 YA were suitable for the analysis of GSC. Independent t-test comparisons indicated that these samples were not significantly different in mean GSC values (p > .05). Thus, to control for the disproportionate sample sizes, GSC means were combined across age groups and were entered into a three-way (Height [Low vs High] x Position [Middle vs Edge] x Task [QS vs BST]) repeated-measures (RM) analysis of variance (ANOVA) using the combined subject sample of 16 participants. Data from BST trials were not analyzed for two female OAs because they were unable to perform the task. Mean durations from the BST task and error rates were analyzed in three-way (Age [YA vs OA] x Height [Low vs High] x Position [Middle vs Edge]) RM analysis of covariance (ANCOVA) using BST values and error rates from the sitting LM condition as a covariate, to control for differences in response times and memory performance (18)(20). Mean COP area data from the same subject sample (15 YA and 13 OA) were entered into a four-way (Age [YA vs OA] x Height [Low vs High] x Position [Middle vs Edge] x Task [QS vs BST]) RM ANOVA. Task was included as a variable to determine the effects of presentation of the BST task on measures of COP.
A prioritization index was calculated to quantify changes in the relationship between postural control and secondary task performance. The prioritization index was obtained by comparing the percentage change in postural performance and secondary task performance across the testing conditions. Change in postural performance was assessed using COP area by calculating the percentage change from the LM condition while performing the BST in the LE, HM, and HE conditions (e.g., [LM – LE]/LM *100). Change in secondary task performance was calculated by the percentage change in the LE, HM, and HE conditions from the duration of the LM condition (e.g., [LM – LE]/LM *100). For the prioritization index, a score was assigned for each participant according to the following criteria: (i) posture prioritized = reduced area and longer duration on BST and (ii) no prioritization = all other possible combinations. The number of participants who revealed postural prioritization versus the number of participants who did not prioritize posture first was assessed by chi-square analysis to determine significant differences in frequencies. Findings for all statistical tests were considered to be significant at 
= 0.05, and Bonferroni adjusted post hoc comparison of means (t tests) were used when appropriate.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
|
Measures of Arousal
A significant main effect for task (F[1,15] = 9.1, p = .009) indicated that performing the BST task significantly increased arousal. Significant main effects also emerged for height (F[1,15] = 19.45, p = .000) and position (F[1,15] = 5.04, p = .041) and revealed that arousal increased from the non-elevated to the elevated positions and from the middle to the edge positions. Comparison of means revealed that GSC values increased by approximately 43% from the LM to the HE conditions in no task (QS) trials and by approximately 35% in task (BST) trials. Interestingly, a 51% increase in GSC values emerged between the LMQS and the HEBST trials. These changes are illustrated in Fig. 3.
|
|
Measure of Postural Control
All participants showed significantly reduced COP area in the elevated compared to the non-elevated conditions (F[1,26] = 11.64, p = .002). A significant main effect did not emerge for position (p > .05), although mean values indicated that COP area was reduced in the edge compared to the middle positions.
OA showed a significantly greater magnitude of sway area than YA (F[1,26] = 12.40, p = .002). A Height x Age interaction revealed that the effect of height differed between YA and OA (F[1,26] = 11.73, p = .002). Follow-up comparison of means indicated that OA showed a significant reduction in COP area from the non-elevated to the elevated conditions (p = .001), while COP area remained unchanged among YA (p > .05; OANon-elevated = 1.45 %BOS2, OAElevated = .802 %BOS2; YANon-elevated = .459 %BOS2 vs YAElevated = .416 %BOS2).
Effects of Secondary Task on Postural Control
The main effect for task was nonsignificant (p > .05), although COP area increased by approximately 50% when the secondary task was introduced (see Table 2 ). A Task x Height interaction, however, indicated that the effect of task differed between heights (F[1,26] = 4.91, p = .036). Post hoc comparisons confirmed that performing the BST task significantly increased COP area in the non-elevated testing conditions (p = .014).
Despite this effect for the BST task in the non-elevated conditions, a significant Task x Height x Age interaction (F[1,26] = 4.53, p = .043) indicated that OA showed a significant reduction in COP area when performing the BST task in the elevated conditions compared to the non-elevated conditions (p = .001; OANon-elevated = 1.95 %BOS2, OAElevated = 0.834 %BOS2; YANon-elevated = 0.505 %BOS2, YAElevated = 0.450 %BOS2; see Fig. 5). This effect did not emerge among YA (p > .05; Fig. 5).
|
level.
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Measures of arousal confirmed that the manipulation of environmental context was successful in increasing arousal. In addition, arousal was found to increase when participants performed the BST task. This finding confirms results from earlier work examining changes in GSC with task performance (21). Interestingly, data indicated that the introduction of the secondary task adversely affected postural control in the non-elevated conditions when postural threat was minimal. Although these findings were not significant, these results are consistent with other research that indicates that the introduction of a secondary task has a negative effect on postural control (7)(22). In the present study, the disruptive effect of concurrent secondary task performance on postural control was particularly evident in OA, among whom COP area measures increased more than 60% under dual task conditions. However, when OA participants were asked to perform the BST task in more threatening conditions, this effect of secondary task performance disappeared. In fact, our findings suggest that postural control among OA improved under dual-task conditions in a threatening environment. We have interpreted this finding to indicate that the prioritization of postural control is dependent upon age and the level of postural threat.
Our results suggest that a reciprocal relationship exists between secondary task performance and postural control among OA; in order for OA to improve postural stability, secondary task performance must deteriorate. Because the BST task has been shown to compete for the same cognitive resources involved in postural control (8)(22), we propose that the concurrent performance of both the postural and cognitive tasks exceeded the cognitive resources available for postural control in OA in the most threatening condition. Indeed, OA have been shown to have a reduced ability to perform multiple tasks because of age-related declines in cognitive capacities (14). According to the model for limited capacity of attention (23), interference from the concurrent performance of a secondary task can affect performance of the primary task, the secondary task, or both. However, a variety of internal or external factors can determine which task is affected. As has been shown in previous studies that employ a dual-task methodology, our results indicated that secondary task performance was maintained at the expense of postural control in nonthreatening conditions (7)(8). It is possible that, as forwarded by "the posture first hypothesis" (7), secondary task performance in our study was prioritized in low threat conditions at the expense of postural control because balance was not perceived to be threatened. However, the unique finding in our study was that, as the consequences of postural instability increased, the relationship between secondary task performance and postural control was altered such that balance was prioritized at the expense of cognitive task performance.
Interestingly, among YA, secondary task performance and measures of postural stability appeared to benefit from conditions of increased arousal. In particular, data from performance of the BST task revealed that performance on the BST task became significantly faster in the edge and in the elevated conditions. We propose that the concurrent performance of the postural and secondary tasks did not exceed the cognitive capacities of YA as it did in OA. It is possible, therefore, that the effects of arousal may have had beneficial effects on secondary task performance in YA. The Yerkes-Dodson law dictates that task performance is improved with increased arousal (24)(25). Thus, according to this theory, YA may have improved performance on the secondary task because of increased arousal. Furthermore, although not significant, our results indicated that postural control was improved in conditions of increased postural threat in YA. Improved postural performance under conditions of threat has been well demonstrated in previous investigations (11)(12). Thus, we believe that the regulation of postural control in response to postural threat is not age-dependent; however, under dual-task conditions, OA may need to sacrifice secondary task performance to modulate postural stability.
The present study has examined the effects of performing a secondary task on postural control in YA and OA when in conditions that increase the consequences of instability. Results indicated that among OA, postural control was significantly improved at the expense of secondary task performance when in conditions of the greatest postural threat. We believe that our work confirms the hypothesis forwarded by Shumway-Cook and colleagues (7) that postural control will be prioritized over secondary task performance when the potential consequences of instability increase.
Understanding how postural control is prioritized may lead to insights regarding how to reduce falls in older adults. Based on our findings, fall risk may increase in OA who are unable, or unwilling, to compromise secondary task performance in conditions of increased postural threat. This finding could have implications for rehabilitation practitioners who often instruct patients while they are performing postural tasks (26). Although our findings contribute to the growing body of knowledge regarding cognition and postural control, it remains unknown how postural control may be altered in individuals with an existing fear of falling. Perhaps individuals with an intense fear of falling may not be able to perform a secondary task in conditions of increased postural threat. Therefore, although the present findings are important in deciphering the relationship between secondary task performance and postural control in different environmental contexts, further research is warranted to determine what actions should be taken to prevent falls in individuals who may be at the greatest risk of falling.
Indeed, the asymmetric gender composition of the YA and OA subject samples may threaten the internal validity of our findings. We believe, however, that the gender composition of the subject samples presented in this study does provide a generalizable representation of demographic trends and, thus, strengthens the external validity of our findings. On a similar note, the loss of participant data in the analysis of GSC results eliminated the possibility for an age comparison regarding GSC effects. An age-related difference in GSC results would strengthen the age-dependent differences regarding the effects of the secondary task on postural control; however, this interpretation cannot be made from our work.
|
Acknowledgments |
|---|
We gratefully acknowledge the contributions of Mitchell Champney, Kendra Massie, and Nicole McKenzie for assistance with data collection, as well as Greg Tompkins and Frank Klassen for equipment construction and hardware technical assistance.
Received February 15, 2002
Accepted June 5, 2002
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. J. St George, R. C. Fitzpatrick, M. W. Rogers, and S. R. Lord Choice Stepping Response and Transfer Times: Effects of Age, Fall Risk, and Secondary Tasks J. Gerontol. A Biol. Sci. Med. Sci., May 1, 2007; 62(5): 537 - 542. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Laufer, Y. Barak, and I. Chemel Age-related differences in the effect of a perceived threat to stability on postural control. J. Gerontol. A Biol. Sci. Med. Sci., May 1, 2006; 61(5): 500 - 504. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
