- Email: [email protected]
Effect of cognitive load on postural control
Brain Research Bulletin, Vol. 58, No. 1, pp. 135–139, 2002 Copyright © 2002 Elsevier Science Inc. All rights reserved. 0361-9230/02/$–see front matter
PII: S0361-9230(02)00770-0
Effect of cognitive load on postural control Gerhard Andersson,1,2∗ Jenni Hagman,1 Roya Talianzadeh,1 Alf Svedberg2 and Hans Christian Larsen2 1 Department
of Psychology, Uppsala University, Uppsala, Sweden; and 2 Department of Audiology, Uppsala University Hospital, Uppsala, Sweden
[Received 21 October 2001; Revised 1 February 2002; Accepted 12 February 2002] consistent, few studies have found that subjects and patients with vestibular dysfunction perform less well on cognitive tasks while their balance is challenged [3,12,14,18,25], with increasing age as an influential factor [22]. Other studies have not replicated this effect [26], and apart from old age and vestibular deficits there is no obvious pattern in the data, which would explain the discrepancies. However, the difficulty of the cognitive task has been proposed as one factor, and also the difficulty of the balance task [14,17,25,28]. For example, Redfern et al. [20] found that the greatest influence postural sway on simple reaction time was found for their old participants when they performed the most difficult balance task (i.e., the sway-referenced visual scene condition). A distinction can be made between verbal reaction time tasks, motor reaction time tasks, and cognitive tests involving a higher working memory load [7,22]. Although it might be assumed that more interference should be consistently found on the performance of more difficult cognitive tasks, this has not always been the case. For example, Dault et al. [8] found no effects of difficulty on three different working memory tasks (a verbal task, a visuo-spatial task with two levels of difficulty, and a central executive task). A ‘posture first’ principle has been suggested that could explain deterioration in cognitive task performance during dual-task studies [22]. This principle would certainly account for the finding that difficult balance tasks result in more cognitive interference, as balance is then prioritised. The second main outcome of interest in cognition–balance dual-task studies is the effect of cognitive tasks on balance functioning. There are several factors that may mediate the effects of cognitive load on balance functioning. The first is attention towards balance [15,16]. A second mediating factor is arousal that could be linked to the performance of both cognitive and postural tasks [4,15]. A third factor is the choice of balance parameter used in the study and the difficulty of the balance task. Recently, it was observed that any task effect must consider the effect of respiration [4,19,26], which precludes the use of verbal cognitive tasks [26]. However, considering this effect, there might still be an effect of silent mental activity on the amount of sway [7,12,15,16,18,22,28]. The literature on the effects of cognitive tasks on balance performance does fairly consistently show that older subjects’ balance is more affected than younger subjects are when they concurrently perform a cognitive task [5,18]. Research on younger participants without balance complaints has been less consistent, and in some studies there has been an absence of an effect [26], or even less sway (stiffness, less sway variability) when performing a cognitive task [8,10]. This observation would
ABSTRACT: The present study reports findings from two experiments on the relation between a mental task (silent backward counting) and posture. The first experiment included 30 normal subjects and the second experiment 20 normal subjects. In Experiment 1 postural sway and performance of the mental task were measured in a 2 × 2 dual-task design (with or without mental task and calf stimulation). In Experiment 2 a similar design was used, the only difference being that during trials without the mental task, subjects were instructed to focus on their balance and provide a rating of body sway. Results showed that balance perturbation led to decreased performance on the cognitive task in Experiment 1, but not in Experiment 2. The mental task led to less body sway, while focused attention attenuated the effect. In conclusion, control of body sway and cognitive functioning are to some extent related. © 2002 Elsevier Science Inc. All rights reserved. KEY WORDS: Posture, Attention, Cognition, Mental task, Vestibular test.
INTRODUCTION The relation between cognition and balance function has received renewed attention recently, with increased recognition of the importance of conscious processes in the regulation of posture [21]. While it is likely that postural adjustments to some extent are automatic and not consciously controlled, more demanding situations call for attentional strategies and monitoring of balance [3,12,14,18,25]. The possible overlap between posture and cognitive functioning has been investigated using dual-task designs in which subjects have performed cognitive and postural tasks simultaneously [26]. The cognitive tasks used have varied from simple reaction time [14,25] to more demanding spatial tasks [3,12]. The postural tasks used in these studies have also differed from standing heel-to-toe (e.g., Romberg task) [12] to computerised posturography, in which balance is perturbed by either visual, somatosensory and vestibular information for controlling balance [19]. In dual-task studies effects must be found when comparing the task (e.g., cognitive task or balance task) in isolation with concurrent performance of another tasks. Any change above the baseline level of performance can be regarded as a dual-task effect with the implication that a competition for central processing resources is taking place. There are two outcomes of interest in dual-task studies on cognition and balance. First, there is the possible effect of balancing on the performance of the cognitive task. Although not
∗ Address for correspondence: Dr. Gerhard Andersson, Department of Psychology, Uppsala University, Box 1225, SE-751 42 Uppsala, Sweden. Fax: +46-18-471-21-23; E-mail: [email protected]
135
136
ANDERSSON ET AL.
be in line with the posture first principle, and could even be interpreted as an overcompensation effect. Thus, when the young adult prioritises balance and focused on stabilisation, less sway (alternatively more stiffness) is observed. As there are few studies on balance–cognition interaction in younger adults [10], the aim of this study was to test the posture first principle by means of a dual-task study. We hypothesised that younger adults would show dual-task decrement on the cognitive task while performing a demanding balance task. Lack of simultaneous decrement on the balance task is interpreted as a shift towards the balance task. This shift, we expected, could even result in decreased sway on the balance task. To test if concentration on balance affects postural sway, we made an attempt to control attention towards balance in second experiment. As attention towards balance is likely to decrease sway (by means of increased arousal), we predicted a weaker dual-task effect on balance in the single- versus dual-task condition. As the performance of the cognitive task in single- and dual-task conditions did not involve any manipulation of attention, we made no prediction of the outcome apart from an expected dual-task decrement. The first experiment reported was designed firstly to assess the effects of a balance task, in which posture was perturbed by stimulation of the calf muscles on the performance of a silent mental arithmetic task (backwards silent counting). Secondly, the effect of the mental task on balance functioning was assessed. In total there were four conditions: (1) standing on a platform with no calf stimulation; (2) standing on a platform with no calf stimulation and simultaneously counting backwards silently; (3) standing on a platform with calf stimulation to perturb balance; (4) standing on a platform with calf stimulation and simultaneously counting backwards silently. A second experiment was conducted which was designed to control for the effects of focused attention towards balance. The same conditions as in Experiment 1 were repeated with the additional instruction to monitor balance and make a rating of postural perturbation on conditions 1 and 3.
delivered at the frequency of 80 Hz (1.0 mm amplitude) with an effect of 1040 mW. Each trial, with or without vibratory stimulation, lasted 20 s with a total of 23 trials pre-programmed. The first three trials were used to acclimatise the subject to the vibration stimuli and subjects had the straps and the vibrators on for the whole experiment. Mental Tasks The mental task condition was silent backward counting in steps of seven as fast and as accurately as possible for 20 s, beginning with randomly selected numbers. Following each trial the subjects were asked their final number which was later checked for accuracy. No feedback on performance was given during the testing proper. Subjects were first instructed how to do the counting backward task and were corrected until it was mastered. In Experiment 2 attention towards balance and the vibratory stimuli was controlled by means of letting the subjects perform a rating task silently while standing on the platform. The task consisted of reporting a rating of how much the balance had been affected while standing on the platform (e.g., each trial). This numerical rating (not at all affected = 0, much affected = 5), was obtained immediately following each trial and was made both following vibratory calf stimulation and on the trials without vibratory stimulation. Design Trials were blocked by condition with four counterbalanced conditions: just standing without vibration, silent counting without vibration, vibration only, and finally vibration and silent counting. Each condition consisted of five trials with a short rest between each block of trials. In Experiment 2 subjects were instructed to direct their attention towards their balance when not doing the silent counting. Blood pressure and heart rate data were obtained before and after the actual testing (Omron, M4, Japan). Because no changes were obtained, these data are not reported. A set of self-report questionnaires was administrated following the test procedures.
MATERIALS AND METHODS Subjects The subjects in Experiment 1 were 30 healthy volunteers (mean age = 27.4 years, SD = 8.15; 17 female) with no history of balance dysfunction or any other central or peripheral dysfunction. Experiment 2 included 20 additional healthy volunteers (mean age = 30.1 years, SD = 8.62; 10 female). Postural sway measurement: body sway was measured by a force platform equipped with strain gauges from which centre of pressure was detected [23]. The system was attached to a IBM compatible computer, and driven by a program which generates a measure of body sway velocity in terms of torque variance ([Nm/mass]2 × 1000) (Postcon, Department of Otorhinolaryngology, University Hospital, Lund, Sweden). Torque variance is proportional to the energy used by the subject to maintain upright posture [11]. Changes in centre point of force actuated by the feet on the platform were digitised and sampled at 10 Hz. Measures of anterior–posterior and lateral sway were obtained. All sway measures were square root transformed before analysis. The subjects were tested while standing on the force platform (height 24 cm; diameter 49×49 cm), facing forwards with extended knees and feet at 30◦ , arms positioned along their sides and with their eyes closed. On half of the trials applying vibratory stimulation to the gastrocnemius muscles of both legs evoked body sway. The vibrators were cylinder shaped (10 cm × 1.5 cm) and held in place by straps around the calf muscles. Vibration stimulation was
RESULTS Mean results on the backward silent counting task are presented in Table 1. In Experiment 1 a dual-task effect was found with paired t-test in that the number of multiples of seven subtracted were lower during the dual-task condition [t (28) = 2.31, p = 0.0283]. In Experiment 2 no statistically significant difference was found between single- and dual-task conditions. In Experiment 1 analyses of the anterior–posterior sway data with a repeated measures ANOVA showed a significant repeated measures effect [F (3, 87) = 53.15, p < 0.0001]. Post-hoc t-tests revealed a significant effect of doing the mental task while receiving vibratory calf stimulation compared to the calf stimulation only [t (29) = 2.15, p = 0.04]. The subjects swayed less when doing the mental task (see Fig. 1). The effect was not significant when TABLE 1 MEAN MULTIPLES OF SEVEN SUBSTRACTED (SDs IN BRACKETS) AS A FUNCTION OF STANDING ON THE PLATFORM ONLY AND DUAL-TASK (VIBRATORY CALF STIMULATION) IN EXPERIMENT 1 (n = 30) AND 2 (n = 20)
Experiment 1 Experiment 2
Counting only
Dual-task
4.21 (2.48) 3.77 (2.58)
3.52 (2.29) 3.9 (2.04)
COGNITIVE LOAD ON POSTURAL CONTROL
FIG. 1. Effects of vibratory calf stimulation and silent backward counting on anterior–posterior sway (Means and SEMs).
comparing silent counting only with standing on the platform without stimulation [t (29) = 1.68, p = 0.10]. The effect of vibratory calf stimulation was evident both without the mental task [t (29) = 9.25, p < 0.0001] and while doing the mental task [t (29) = 7.47, p < 0.0001]. Subjects swayed more when stimulated. In Experiment 2 the effect of silent counting was compared with focused attention towards the balance. Mean rating of how balance was affected (pooled across trials within conditions) showed a mean rating of M = 3.69 (SD = 1.21) following calf stimulation and a M = 1.49 (SD = 0.67) following just standing on the platform. The difference was significant by means of paired t-test [t (19) = 7.82, p < 0.0001]. Analyses of the sway data in Experiment 2 (see Fig. 1) with ANOVA showed a significant repeated measures effect [F (3, 57) = 13.6, p < 0.0001]. Post-hoc t-tests showed no significant differences between doing and not doing the mental task, while the effect of vibratory calf stimulation was found both when focusing attention [t (19) = 4.3, p = 0.004] and when doing the mental task [t (19) = 3.78, p < 0.001]. Analysis of lateral sway data in Experiment 1 (Fig. 2) showed a repeated measures effect [F (3, 87) = 11.73, p < 0.0001]. Post-hoc t-tests showed a significant effect of doing the mental task while receiving vibratory calf stimulation compared to the calf stimulation only [t (29) = 2.10, p = 0.045]. Post-hoc t-tests also showed a significant effect of doing the mental task when not receiving vibratory calf stimulation compared to standing on the platform only [t (29) = 2.60, p = 0.014]. Subjects swayed less while doing the mental task. Again the predicted destabilising effect of vibratory stimulation was evident in both conditions, both with [t (29) = 4.58, p < 0.0001] and without the mental task [t (29) = 2.7, p = 0.012]. In Experiment 2 analysis of lateral sway again resulted in a repeated measures effect [F (3, 57) = 7.35, p = 0.0003]. Post-hoc t-tests indicated that the subjects swayed less while doing the silent
137
FIG. 2. Effects of vibratory calf stimulation and silent backward counting on lateral sway (Means and SEMs).
counting + stimulation compared with just receiving stimulation [t (19) = 2.06, p = 0.054]. The effects of calf stimulation were significant as predicted [t (19) = 3.36, p = 0.003 and t (19) = 3.4, p = 0.004], with or without doing the mental task. DISCUSSION The findings in this study indicate that control of body sway and cognitive functioning are not independent systems, and give some additional support to the posture first principle. Our first purpose was to study the effect of balance perturbation on the performance of a cognitive task. Whereas a dual-task effect was found in Experiment 1, with lower scores obtained when the balance was challenged, this effect was not observed in Experiment 2. Decrements in the performance of mental tasks have been observed in other studies [3,12,14,18,25], but there is no obvious procedural variation that would easily explain the discrepancy in results between Experiments 1 and 2. One difference was the sample size, which was smaller in Experiment 2, but inspection of the means in Table 1 shows that this is not a plausible explanation as the means were close to identical. Moreover, in Experiment 2 attention was manipulated during the conditions when the subjects did not perform the cognitive task, but during the cognitive task conditions were the same in Experiments 1 and 2. Future studies could investigate the effects on a selection of different cognitive tasks, as these may not show similar results, or even opposite results. An example is Maylor and Wing [18] who found impaired performance on random generation of numbers when balance was disturbed. An opposite effect was found for silent counting in which subjects required less time to do under dual-task conditions. Silent backward counting might less sensitive to dual-task effects than reaction time tasks. Supporting this notion, Yardley et al. [26] found no dual-task effect
138 using the same cognitive task. However, again with the same cognitive task, dual-task decrement has been found following passive rotation in the dark in healthy subjects [27], and as a result of a concurrent orientation monitoring task in both healthy and balance impaired subjects [29]. Whereas a few earlier studies had indicated increased instability by spoken mental tasks [3,26], the present study used silent backward counting and found a tendency for the opposite. In one of the conditions in the experiment by Yardley et al. [26] a similar procedure as the one presented here was used. Although the effect of backward silent counting on sway failed to reach statistical significance, it was found that their subjects swayed slightly less when counting silently backwards than when no mental task was performed. In a more recent study, decreased sway during the performance of a mental task was found in a stable platform condition [28]. In the present study an effect of less sway when doing a similar mental task was also confirmed statistically. However, a reservation must be made that our postural sway measures were limited and that other alternative parameters (e.g., centre of mass variability) might yield other results. In the study balance was perturbed and as expected subjects swayed more when the vibratory calf stimulation was applied [2]. Interestingly, we found an effect of cognitive tasking on sway both with and without vibratory calf stimulation (in four out of eight possible instances). If any, the effect of focused attention is to decrease the difference in sway between counting backwards and just standing on the platform. An alternative hypothesis is that the effect is mediated by arousal. Both the backward silent counting and the balance rating (focused attention) tasks may have increased arousal in comparison with the baseline task of standing without performing any task. A link between arousal and postural strategies has been proposed [15,16], but the causal links have not been investigated in depth. From a theoretical point of view, it is plausible that the arousal and stress caused by a mental task will cause persons to sharpen their balance. This assumption would only be valid if decreased sway (here defined as torque variance), would be functionally linked to better balance. Given this assumption, decreases in sway when faced with stress would likely be a highly adaptive response. However, research by Stoffregen et al. [24] has questioned the posture first principle, and instead they do not regard stability as an end on its own right, but rather as a means to an end (for example, preparation or execution of movement). We concur with this ecologically valid assumption, but believe that it is compatible with the posture first principle, given that decreased sway (if equal to better balance) could be adaptive and serve as a means to a goal. In the analysis by Stoffregen et al. [24] they suggested that upright stance is not optimal to avoid the risk of falling, and that sitting or lying down would be better. However, as they saw a role of suprapostural behaviours, it could be the case that the role of upright stance (in good balance) would be to facilitate rapid movement. A related hypothesis is that fear will cause persons to sharpen their balance in terms of displaying less sway, or stiffening [1,6] as a response to balance perturbation. This would make sense from an evolutionary point of view and continued research is under way to investigate this possibility by administrating aversive stimulation (e.g., noise bursts) while balance is measured. While there are several studies on fear of falling in the elderly [13], to our knowledge less is known about the influence of emotion on sway, using experimental approaches. One recent exception is a study by Adkin et al. [1] who found a progressively tightened control of posture as the level of postural threat increased (induced by alterations of surface height). The effects of cognitive workload on balance have not been researched in depth. For example, it could be that more demanding
ANDERSSON ET AL. cognitive tasks than backwards silent counting led to decreases of sway, while still being receptive to the effect of balance perturbation on cognitive performance. Scaling procedures with increasingly difficult balance and cognitive tasks could be used in future dual-task studies on the interference between posture and cognition. Finally, the results should be extended to clinical samples of patients with vestibular disorders in which anxiety and fear are common features [9]. CONCLUSION Findings from this study suggest that the performance of a silent mental arithmetic task (counting backwards in multiples of seven) is impaired when balance is perturbed. Instead of increasing sway, counting backwards silently led to a decrease in sway. Conversely, focused attention on balance had a small effect, and attenuated the effect of the cognitive task. The inconsistencies in the reported literature on cognition–balance interference may be caused by the use of cognitive tasks of varying complexity, as well as differing balance tasks. In order to assess the role of cognition in the control of balance it is important to investigate the circumstances under which balance is affected by concurrent mental tasks and in particular to investigate the role of emotion. ACKNOWLEDGEMENTS
This research was supported by a project grant from the Swedish Council for Research in the Humanities and Social Sciences. REFERENCES 1. Adkin, A. L.; Frank, J. S.; Carpenter, M. G.; Peysar, G. W. Postural control is scaled to level of postural threat. Gait Posture 14:87–93; 2000. 2. Andersson, G.; Persson, K.; Melin, L.; Larsen, H.-C. Actual and perceived postural sway during balance specific and non-specific proprioceptive stimulation. Acta Otolaryngol. (Stockholm) 118:461– 465; 1998. 3. Andersson, G.; Yardley, L.; Luxon, L. A dual task study of interference between mental activity and control of balance. Am. J. Otolaryngol. 19:632–637; 1998. 4. Bouisset, S.; Duchéné, J.-L. Is body balance more perturbed by respiration in seating than in standing posture? Neuroreport 5:957– 960; 1994. 5. Brauer, S. G.; Woollacott, M.; Shumway-Cook, A. The interacting effects of cognitive demand and recovery of postural stability in balance-impaired elderly persons. J. Gerontol. Med. Sci. 56A:M489– M496; 2001. 6. Carpenter, M. G.; Frank, J. S.; Silcher, C. P. Surface height effects on postural control: A hypothesis for a stiffness strategy for stance. J. Vestib. Res. 9:277–286; 1999. 7. Dault, M. C.; Frank, J. S.; Allard, F. Influence of a visuo-spatial, verbal and central executive working memory task on postural control. Gait Posture 14:110–116; 2001. 8. Dault, M. C.; Geurts, A. C. H.; Mulder, T. W.; Duysens, J. Postural control and cognitive task performance in healthy participants while balancing on different support-surface configurations. Gait Posture 14:248–255; 2001. 9. Furman, J. M.; Jacob, R. G. Psychiatric dizziness. Neurology 48:1161– 1166; 1997. 10. Hunter, M. C.; Hoffman, M. A. Postural control: Visual and cognitive manipulations. Gait Posture 13:41–48; 2001. 11. Johansson, R.; Magnusson, M. Human postural dynamics. CRC Crit. Rev. Biomed. Eng. 18:413–437; 1991. 12. Kerr, B.; Condon, S. M.; McDonald, L. A. Cognitive spatial processing and the regulation of posture. J. Exp. Psychol. 11:617–622; 1985. 13. Lachman, M. E.; Howland, J.; Tennstedt, S.; Jette, A.; Assman, S.; Peterson, E. W. Fear of falling and activity restriction: The survey of activities and fear of falling in the elderly (SAFE). J. Gerontol. Psych. Sci. 53B:P43–P50; 1998.
COGNITIVE LOAD ON POSTURAL CONTROL 14. Lajoie, Y.; Teasdale, N.; Bard, C.; Fleury, M. Attentional demands for static and dynamic equilibrium. Exp. Brain Res. 97:139–144; 1993. 15. Maki, B. E.; Whitelaw, R. S. Influence of expectation and arousal on center-of-pressure responses to transient postural perturbations. J. Vestib. Res. 3:25–39; 1993. 16. Maki, B. E.; McLlroy, W. E. Influence of arousal and attention on the control of postural sway. J. Vestib. Res. 6:53–59; 1996. 17. Marsh, A. P.; Geel, S. E. The effect of age on the attentional demands of postural control. Gait Posture 12:105–113; 2000. 18. Maylor, E. A.; Wing, A. M. Age differences in postural stability are increased by additional cognitive demands. J. Gerontol. Psych. Sci. 51:P143–P154; 1996. 19. Nasher, L. M.; Peters, J. F. Dynamic posturography in the diagnosis and management of dizziness and balance disorders. Diagnostic Neurol. 8:331–349; 1990. 20. Redfern, M. S.; Jennings, J. R.; Martin, C.; Furman, J. M. Attention influences sensory integration for postural control in older adults. Gait Posture 14:211–216; 2001. 21. Schlesinger, A.; Redfern, M. S.; Dahl, R. E.; Jennings, J. R. Postural control attention and sleep deprivation. Neuroreport 9:49–52; 1998. 22. Shumway-Cook, A.; Woollacott, M.; Kerns, K. A.; Baldwin, M. The effects of two types of cognitive tasks on postural stability in older adults with and without a history of falls. J. Gerontol. Med. Sci. 52:M232–M240; 1997.
139 23. Starck, J.; Pyykkö, I.; Aalto, H.; Pekkarinen, J. Measurement of postural stability: Development of a force platform and some excitation systems. Med. Prog. Technol. 18:209–215; 1993. 24. Stoffregen, T. A.; Smart, L. J.; Bardy, B. G.; Pagulayan, R. J. Postural stabilization of looking. J. Exp. Psychol.: Learning, Memory, Cognition 25:1641–1658; 1999. 25. Teasdale, N.; Bard, C.; LaRue, J.; Fleury, M. On the cognitive penetrability of posture control. Exp. Aging Res. 19:1–13; 1993. 26. Yardley, L.; Gardner, M.; Leadbetter, A.; Lavie, N. Effect of articulatory and mental tasks on postural control. Neuroreport 10:215– 219; 1999. 27. Yardley, L.; Gardner, M.; Lavie, N.; Gresty, M. Attentional demands of perception of passive self-motion in darkness. Neuropsychologia 37:1293–1301; 1999. 28. Yardley, L.; Gardner, M.; Bronstein, A.; Davies, R.; Buckwell, D.; Luxon, L. Interference between postural control and mental task performance in patients with vestibular disorder and healthy controls. J. Neurol. Neurosurg. Psychiatr. 71:48–52; 2001. 29. Yardley, L.; Papo, D.; Bronstein, A.; Gresty, M.; Gardner, M.; Lavie, N.; Luxon, L. Attentional demands of continuously monitoring orientation using vestibular information. Neuropsychologia 40:373– 383; 2002.
PII: S0361-9230(02)00770-0
Effect of cognitive load on postural control Gerhard Andersson,1,2∗ Jenni Hagman,1 Roya Talianzadeh,1 Alf Svedberg2 and Hans Christian Larsen2 1 Department
of Psychology, Uppsala University, Uppsala, Sweden; and 2 Department of Audiology, Uppsala University Hospital, Uppsala, Sweden
[Received 21 October 2001; Revised 1 February 2002; Accepted 12 February 2002] consistent, few studies have found that subjects and patients with vestibular dysfunction perform less well on cognitive tasks while their balance is challenged [3,12,14,18,25], with increasing age as an influential factor [22]. Other studies have not replicated this effect [26], and apart from old age and vestibular deficits there is no obvious pattern in the data, which would explain the discrepancies. However, the difficulty of the cognitive task has been proposed as one factor, and also the difficulty of the balance task [14,17,25,28]. For example, Redfern et al. [20] found that the greatest influence postural sway on simple reaction time was found for their old participants when they performed the most difficult balance task (i.e., the sway-referenced visual scene condition). A distinction can be made between verbal reaction time tasks, motor reaction time tasks, and cognitive tests involving a higher working memory load [7,22]. Although it might be assumed that more interference should be consistently found on the performance of more difficult cognitive tasks, this has not always been the case. For example, Dault et al. [8] found no effects of difficulty on three different working memory tasks (a verbal task, a visuo-spatial task with two levels of difficulty, and a central executive task). A ‘posture first’ principle has been suggested that could explain deterioration in cognitive task performance during dual-task studies [22]. This principle would certainly account for the finding that difficult balance tasks result in more cognitive interference, as balance is then prioritised. The second main outcome of interest in cognition–balance dual-task studies is the effect of cognitive tasks on balance functioning. There are several factors that may mediate the effects of cognitive load on balance functioning. The first is attention towards balance [15,16]. A second mediating factor is arousal that could be linked to the performance of both cognitive and postural tasks [4,15]. A third factor is the choice of balance parameter used in the study and the difficulty of the balance task. Recently, it was observed that any task effect must consider the effect of respiration [4,19,26], which precludes the use of verbal cognitive tasks [26]. However, considering this effect, there might still be an effect of silent mental activity on the amount of sway [7,12,15,16,18,22,28]. The literature on the effects of cognitive tasks on balance performance does fairly consistently show that older subjects’ balance is more affected than younger subjects are when they concurrently perform a cognitive task [5,18]. Research on younger participants without balance complaints has been less consistent, and in some studies there has been an absence of an effect [26], or even less sway (stiffness, less sway variability) when performing a cognitive task [8,10]. This observation would
ABSTRACT: The present study reports findings from two experiments on the relation between a mental task (silent backward counting) and posture. The first experiment included 30 normal subjects and the second experiment 20 normal subjects. In Experiment 1 postural sway and performance of the mental task were measured in a 2 × 2 dual-task design (with or without mental task and calf stimulation). In Experiment 2 a similar design was used, the only difference being that during trials without the mental task, subjects were instructed to focus on their balance and provide a rating of body sway. Results showed that balance perturbation led to decreased performance on the cognitive task in Experiment 1, but not in Experiment 2. The mental task led to less body sway, while focused attention attenuated the effect. In conclusion, control of body sway and cognitive functioning are to some extent related. © 2002 Elsevier Science Inc. All rights reserved. KEY WORDS: Posture, Attention, Cognition, Mental task, Vestibular test.
INTRODUCTION The relation between cognition and balance function has received renewed attention recently, with increased recognition of the importance of conscious processes in the regulation of posture [21]. While it is likely that postural adjustments to some extent are automatic and not consciously controlled, more demanding situations call for attentional strategies and monitoring of balance [3,12,14,18,25]. The possible overlap between posture and cognitive functioning has been investigated using dual-task designs in which subjects have performed cognitive and postural tasks simultaneously [26]. The cognitive tasks used have varied from simple reaction time [14,25] to more demanding spatial tasks [3,12]. The postural tasks used in these studies have also differed from standing heel-to-toe (e.g., Romberg task) [12] to computerised posturography, in which balance is perturbed by either visual, somatosensory and vestibular information for controlling balance [19]. In dual-task studies effects must be found when comparing the task (e.g., cognitive task or balance task) in isolation with concurrent performance of another tasks. Any change above the baseline level of performance can be regarded as a dual-task effect with the implication that a competition for central processing resources is taking place. There are two outcomes of interest in dual-task studies on cognition and balance. First, there is the possible effect of balancing on the performance of the cognitive task. Although not
∗ Address for correspondence: Dr. Gerhard Andersson, Department of Psychology, Uppsala University, Box 1225, SE-751 42 Uppsala, Sweden. Fax: +46-18-471-21-23; E-mail: [email protected]
135
136
ANDERSSON ET AL.
be in line with the posture first principle, and could even be interpreted as an overcompensation effect. Thus, when the young adult prioritises balance and focused on stabilisation, less sway (alternatively more stiffness) is observed. As there are few studies on balance–cognition interaction in younger adults [10], the aim of this study was to test the posture first principle by means of a dual-task study. We hypothesised that younger adults would show dual-task decrement on the cognitive task while performing a demanding balance task. Lack of simultaneous decrement on the balance task is interpreted as a shift towards the balance task. This shift, we expected, could even result in decreased sway on the balance task. To test if concentration on balance affects postural sway, we made an attempt to control attention towards balance in second experiment. As attention towards balance is likely to decrease sway (by means of increased arousal), we predicted a weaker dual-task effect on balance in the single- versus dual-task condition. As the performance of the cognitive task in single- and dual-task conditions did not involve any manipulation of attention, we made no prediction of the outcome apart from an expected dual-task decrement. The first experiment reported was designed firstly to assess the effects of a balance task, in which posture was perturbed by stimulation of the calf muscles on the performance of a silent mental arithmetic task (backwards silent counting). Secondly, the effect of the mental task on balance functioning was assessed. In total there were four conditions: (1) standing on a platform with no calf stimulation; (2) standing on a platform with no calf stimulation and simultaneously counting backwards silently; (3) standing on a platform with calf stimulation to perturb balance; (4) standing on a platform with calf stimulation and simultaneously counting backwards silently. A second experiment was conducted which was designed to control for the effects of focused attention towards balance. The same conditions as in Experiment 1 were repeated with the additional instruction to monitor balance and make a rating of postural perturbation on conditions 1 and 3.
delivered at the frequency of 80 Hz (1.0 mm amplitude) with an effect of 1040 mW. Each trial, with or without vibratory stimulation, lasted 20 s with a total of 23 trials pre-programmed. The first three trials were used to acclimatise the subject to the vibration stimuli and subjects had the straps and the vibrators on for the whole experiment. Mental Tasks The mental task condition was silent backward counting in steps of seven as fast and as accurately as possible for 20 s, beginning with randomly selected numbers. Following each trial the subjects were asked their final number which was later checked for accuracy. No feedback on performance was given during the testing proper. Subjects were first instructed how to do the counting backward task and were corrected until it was mastered. In Experiment 2 attention towards balance and the vibratory stimuli was controlled by means of letting the subjects perform a rating task silently while standing on the platform. The task consisted of reporting a rating of how much the balance had been affected while standing on the platform (e.g., each trial). This numerical rating (not at all affected = 0, much affected = 5), was obtained immediately following each trial and was made both following vibratory calf stimulation and on the trials without vibratory stimulation. Design Trials were blocked by condition with four counterbalanced conditions: just standing without vibration, silent counting without vibration, vibration only, and finally vibration and silent counting. Each condition consisted of five trials with a short rest between each block of trials. In Experiment 2 subjects were instructed to direct their attention towards their balance when not doing the silent counting. Blood pressure and heart rate data were obtained before and after the actual testing (Omron, M4, Japan). Because no changes were obtained, these data are not reported. A set of self-report questionnaires was administrated following the test procedures.
MATERIALS AND METHODS Subjects The subjects in Experiment 1 were 30 healthy volunteers (mean age = 27.4 years, SD = 8.15; 17 female) with no history of balance dysfunction or any other central or peripheral dysfunction. Experiment 2 included 20 additional healthy volunteers (mean age = 30.1 years, SD = 8.62; 10 female). Postural sway measurement: body sway was measured by a force platform equipped with strain gauges from which centre of pressure was detected [23]. The system was attached to a IBM compatible computer, and driven by a program which generates a measure of body sway velocity in terms of torque variance ([Nm/mass]2 × 1000) (Postcon, Department of Otorhinolaryngology, University Hospital, Lund, Sweden). Torque variance is proportional to the energy used by the subject to maintain upright posture [11]. Changes in centre point of force actuated by the feet on the platform were digitised and sampled at 10 Hz. Measures of anterior–posterior and lateral sway were obtained. All sway measures were square root transformed before analysis. The subjects were tested while standing on the force platform (height 24 cm; diameter 49×49 cm), facing forwards with extended knees and feet at 30◦ , arms positioned along their sides and with their eyes closed. On half of the trials applying vibratory stimulation to the gastrocnemius muscles of both legs evoked body sway. The vibrators were cylinder shaped (10 cm × 1.5 cm) and held in place by straps around the calf muscles. Vibration stimulation was
RESULTS Mean results on the backward silent counting task are presented in Table 1. In Experiment 1 a dual-task effect was found with paired t-test in that the number of multiples of seven subtracted were lower during the dual-task condition [t (28) = 2.31, p = 0.0283]. In Experiment 2 no statistically significant difference was found between single- and dual-task conditions. In Experiment 1 analyses of the anterior–posterior sway data with a repeated measures ANOVA showed a significant repeated measures effect [F (3, 87) = 53.15, p < 0.0001]. Post-hoc t-tests revealed a significant effect of doing the mental task while receiving vibratory calf stimulation compared to the calf stimulation only [t (29) = 2.15, p = 0.04]. The subjects swayed less when doing the mental task (see Fig. 1). The effect was not significant when TABLE 1 MEAN MULTIPLES OF SEVEN SUBSTRACTED (SDs IN BRACKETS) AS A FUNCTION OF STANDING ON THE PLATFORM ONLY AND DUAL-TASK (VIBRATORY CALF STIMULATION) IN EXPERIMENT 1 (n = 30) AND 2 (n = 20)
Experiment 1 Experiment 2
Counting only
Dual-task
4.21 (2.48) 3.77 (2.58)
3.52 (2.29) 3.9 (2.04)
COGNITIVE LOAD ON POSTURAL CONTROL
FIG. 1. Effects of vibratory calf stimulation and silent backward counting on anterior–posterior sway (Means and SEMs).
comparing silent counting only with standing on the platform without stimulation [t (29) = 1.68, p = 0.10]. The effect of vibratory calf stimulation was evident both without the mental task [t (29) = 9.25, p < 0.0001] and while doing the mental task [t (29) = 7.47, p < 0.0001]. Subjects swayed more when stimulated. In Experiment 2 the effect of silent counting was compared with focused attention towards the balance. Mean rating of how balance was affected (pooled across trials within conditions) showed a mean rating of M = 3.69 (SD = 1.21) following calf stimulation and a M = 1.49 (SD = 0.67) following just standing on the platform. The difference was significant by means of paired t-test [t (19) = 7.82, p < 0.0001]. Analyses of the sway data in Experiment 2 (see Fig. 1) with ANOVA showed a significant repeated measures effect [F (3, 57) = 13.6, p < 0.0001]. Post-hoc t-tests showed no significant differences between doing and not doing the mental task, while the effect of vibratory calf stimulation was found both when focusing attention [t (19) = 4.3, p = 0.004] and when doing the mental task [t (19) = 3.78, p < 0.001]. Analysis of lateral sway data in Experiment 1 (Fig. 2) showed a repeated measures effect [F (3, 87) = 11.73, p < 0.0001]. Post-hoc t-tests showed a significant effect of doing the mental task while receiving vibratory calf stimulation compared to the calf stimulation only [t (29) = 2.10, p = 0.045]. Post-hoc t-tests also showed a significant effect of doing the mental task when not receiving vibratory calf stimulation compared to standing on the platform only [t (29) = 2.60, p = 0.014]. Subjects swayed less while doing the mental task. Again the predicted destabilising effect of vibratory stimulation was evident in both conditions, both with [t (29) = 4.58, p < 0.0001] and without the mental task [t (29) = 2.7, p = 0.012]. In Experiment 2 analysis of lateral sway again resulted in a repeated measures effect [F (3, 57) = 7.35, p = 0.0003]. Post-hoc t-tests indicated that the subjects swayed less while doing the silent
137
FIG. 2. Effects of vibratory calf stimulation and silent backward counting on lateral sway (Means and SEMs).
counting + stimulation compared with just receiving stimulation [t (19) = 2.06, p = 0.054]. The effects of calf stimulation were significant as predicted [t (19) = 3.36, p = 0.003 and t (19) = 3.4, p = 0.004], with or without doing the mental task. DISCUSSION The findings in this study indicate that control of body sway and cognitive functioning are not independent systems, and give some additional support to the posture first principle. Our first purpose was to study the effect of balance perturbation on the performance of a cognitive task. Whereas a dual-task effect was found in Experiment 1, with lower scores obtained when the balance was challenged, this effect was not observed in Experiment 2. Decrements in the performance of mental tasks have been observed in other studies [3,12,14,18,25], but there is no obvious procedural variation that would easily explain the discrepancy in results between Experiments 1 and 2. One difference was the sample size, which was smaller in Experiment 2, but inspection of the means in Table 1 shows that this is not a plausible explanation as the means were close to identical. Moreover, in Experiment 2 attention was manipulated during the conditions when the subjects did not perform the cognitive task, but during the cognitive task conditions were the same in Experiments 1 and 2. Future studies could investigate the effects on a selection of different cognitive tasks, as these may not show similar results, or even opposite results. An example is Maylor and Wing [18] who found impaired performance on random generation of numbers when balance was disturbed. An opposite effect was found for silent counting in which subjects required less time to do under dual-task conditions. Silent backward counting might less sensitive to dual-task effects than reaction time tasks. Supporting this notion, Yardley et al. [26] found no dual-task effect
138 using the same cognitive task. However, again with the same cognitive task, dual-task decrement has been found following passive rotation in the dark in healthy subjects [27], and as a result of a concurrent orientation monitoring task in both healthy and balance impaired subjects [29]. Whereas a few earlier studies had indicated increased instability by spoken mental tasks [3,26], the present study used silent backward counting and found a tendency for the opposite. In one of the conditions in the experiment by Yardley et al. [26] a similar procedure as the one presented here was used. Although the effect of backward silent counting on sway failed to reach statistical significance, it was found that their subjects swayed slightly less when counting silently backwards than when no mental task was performed. In a more recent study, decreased sway during the performance of a mental task was found in a stable platform condition [28]. In the present study an effect of less sway when doing a similar mental task was also confirmed statistically. However, a reservation must be made that our postural sway measures were limited and that other alternative parameters (e.g., centre of mass variability) might yield other results. In the study balance was perturbed and as expected subjects swayed more when the vibratory calf stimulation was applied [2]. Interestingly, we found an effect of cognitive tasking on sway both with and without vibratory calf stimulation (in four out of eight possible instances). If any, the effect of focused attention is to decrease the difference in sway between counting backwards and just standing on the platform. An alternative hypothesis is that the effect is mediated by arousal. Both the backward silent counting and the balance rating (focused attention) tasks may have increased arousal in comparison with the baseline task of standing without performing any task. A link between arousal and postural strategies has been proposed [15,16], but the causal links have not been investigated in depth. From a theoretical point of view, it is plausible that the arousal and stress caused by a mental task will cause persons to sharpen their balance. This assumption would only be valid if decreased sway (here defined as torque variance), would be functionally linked to better balance. Given this assumption, decreases in sway when faced with stress would likely be a highly adaptive response. However, research by Stoffregen et al. [24] has questioned the posture first principle, and instead they do not regard stability as an end on its own right, but rather as a means to an end (for example, preparation or execution of movement). We concur with this ecologically valid assumption, but believe that it is compatible with the posture first principle, given that decreased sway (if equal to better balance) could be adaptive and serve as a means to a goal. In the analysis by Stoffregen et al. [24] they suggested that upright stance is not optimal to avoid the risk of falling, and that sitting or lying down would be better. However, as they saw a role of suprapostural behaviours, it could be the case that the role of upright stance (in good balance) would be to facilitate rapid movement. A related hypothesis is that fear will cause persons to sharpen their balance in terms of displaying less sway, or stiffening [1,6] as a response to balance perturbation. This would make sense from an evolutionary point of view and continued research is under way to investigate this possibility by administrating aversive stimulation (e.g., noise bursts) while balance is measured. While there are several studies on fear of falling in the elderly [13], to our knowledge less is known about the influence of emotion on sway, using experimental approaches. One recent exception is a study by Adkin et al. [1] who found a progressively tightened control of posture as the level of postural threat increased (induced by alterations of surface height). The effects of cognitive workload on balance have not been researched in depth. For example, it could be that more demanding
ANDERSSON ET AL. cognitive tasks than backwards silent counting led to decreases of sway, while still being receptive to the effect of balance perturbation on cognitive performance. Scaling procedures with increasingly difficult balance and cognitive tasks could be used in future dual-task studies on the interference between posture and cognition. Finally, the results should be extended to clinical samples of patients with vestibular disorders in which anxiety and fear are common features [9]. CONCLUSION Findings from this study suggest that the performance of a silent mental arithmetic task (counting backwards in multiples of seven) is impaired when balance is perturbed. Instead of increasing sway, counting backwards silently led to a decrease in sway. Conversely, focused attention on balance had a small effect, and attenuated the effect of the cognitive task. The inconsistencies in the reported literature on cognition–balance interference may be caused by the use of cognitive tasks of varying complexity, as well as differing balance tasks. In order to assess the role of cognition in the control of balance it is important to investigate the circumstances under which balance is affected by concurrent mental tasks and in particular to investigate the role of emotion. ACKNOWLEDGEMENTS
This research was supported by a project grant from the Swedish Council for Research in the Humanities and Social Sciences. REFERENCES 1. Adkin, A. L.; Frank, J. S.; Carpenter, M. G.; Peysar, G. W. Postural control is scaled to level of postural threat. Gait Posture 14:87–93; 2000. 2. Andersson, G.; Persson, K.; Melin, L.; Larsen, H.-C. Actual and perceived postural sway during balance specific and non-specific proprioceptive stimulation. Acta Otolaryngol. (Stockholm) 118:461– 465; 1998. 3. Andersson, G.; Yardley, L.; Luxon, L. A dual task study of interference between mental activity and control of balance. Am. J. Otolaryngol. 19:632–637; 1998. 4. Bouisset, S.; Duchéné, J.-L. Is body balance more perturbed by respiration in seating than in standing posture? Neuroreport 5:957– 960; 1994. 5. Brauer, S. G.; Woollacott, M.; Shumway-Cook, A. The interacting effects of cognitive demand and recovery of postural stability in balance-impaired elderly persons. J. Gerontol. Med. Sci. 56A:M489– M496; 2001. 6. Carpenter, M. G.; Frank, J. S.; Silcher, C. P. Surface height effects on postural control: A hypothesis for a stiffness strategy for stance. J. Vestib. Res. 9:277–286; 1999. 7. Dault, M. C.; Frank, J. S.; Allard, F. Influence of a visuo-spatial, verbal and central executive working memory task on postural control. Gait Posture 14:110–116; 2001. 8. Dault, M. C.; Geurts, A. C. H.; Mulder, T. W.; Duysens, J. Postural control and cognitive task performance in healthy participants while balancing on different support-surface configurations. Gait Posture 14:248–255; 2001. 9. Furman, J. M.; Jacob, R. G. Psychiatric dizziness. Neurology 48:1161– 1166; 1997. 10. Hunter, M. C.; Hoffman, M. A. Postural control: Visual and cognitive manipulations. Gait Posture 13:41–48; 2001. 11. Johansson, R.; Magnusson, M. Human postural dynamics. CRC Crit. Rev. Biomed. Eng. 18:413–437; 1991. 12. Kerr, B.; Condon, S. M.; McDonald, L. A. Cognitive spatial processing and the regulation of posture. J. Exp. Psychol. 11:617–622; 1985. 13. Lachman, M. E.; Howland, J.; Tennstedt, S.; Jette, A.; Assman, S.; Peterson, E. W. Fear of falling and activity restriction: The survey of activities and fear of falling in the elderly (SAFE). J. Gerontol. Psych. Sci. 53B:P43–P50; 1998.
COGNITIVE LOAD ON POSTURAL CONTROL 14. Lajoie, Y.; Teasdale, N.; Bard, C.; Fleury, M. Attentional demands for static and dynamic equilibrium. Exp. Brain Res. 97:139–144; 1993. 15. Maki, B. E.; Whitelaw, R. S. Influence of expectation and arousal on center-of-pressure responses to transient postural perturbations. J. Vestib. Res. 3:25–39; 1993. 16. Maki, B. E.; McLlroy, W. E. Influence of arousal and attention on the control of postural sway. J. Vestib. Res. 6:53–59; 1996. 17. Marsh, A. P.; Geel, S. E. The effect of age on the attentional demands of postural control. Gait Posture 12:105–113; 2000. 18. Maylor, E. A.; Wing, A. M. Age differences in postural stability are increased by additional cognitive demands. J. Gerontol. Psych. Sci. 51:P143–P154; 1996. 19. Nasher, L. M.; Peters, J. F. Dynamic posturography in the diagnosis and management of dizziness and balance disorders. Diagnostic Neurol. 8:331–349; 1990. 20. Redfern, M. S.; Jennings, J. R.; Martin, C.; Furman, J. M. Attention influences sensory integration for postural control in older adults. Gait Posture 14:211–216; 2001. 21. Schlesinger, A.; Redfern, M. S.; Dahl, R. E.; Jennings, J. R. Postural control attention and sleep deprivation. Neuroreport 9:49–52; 1998. 22. Shumway-Cook, A.; Woollacott, M.; Kerns, K. A.; Baldwin, M. The effects of two types of cognitive tasks on postural stability in older adults with and without a history of falls. J. Gerontol. Med. Sci. 52:M232–M240; 1997.
139 23. Starck, J.; Pyykkö, I.; Aalto, H.; Pekkarinen, J. Measurement of postural stability: Development of a force platform and some excitation systems. Med. Prog. Technol. 18:209–215; 1993. 24. Stoffregen, T. A.; Smart, L. J.; Bardy, B. G.; Pagulayan, R. J. Postural stabilization of looking. J. Exp. Psychol.: Learning, Memory, Cognition 25:1641–1658; 1999. 25. Teasdale, N.; Bard, C.; LaRue, J.; Fleury, M. On the cognitive penetrability of posture control. Exp. Aging Res. 19:1–13; 1993. 26. Yardley, L.; Gardner, M.; Leadbetter, A.; Lavie, N. Effect of articulatory and mental tasks on postural control. Neuroreport 10:215– 219; 1999. 27. Yardley, L.; Gardner, M.; Lavie, N.; Gresty, M. Attentional demands of perception of passive self-motion in darkness. Neuropsychologia 37:1293–1301; 1999. 28. Yardley, L.; Gardner, M.; Bronstein, A.; Davies, R.; Buckwell, D.; Luxon, L. Interference between postural control and mental task performance in patients with vestibular disorder and healthy controls. J. Neurol. Neurosurg. Psychiatr. 71:48–52; 2001. 29. Yardley, L.; Papo, D.; Bronstein, A.; Gresty, M.; Gardner, M.; Lavie, N.; Luxon, L. Attentional demands of continuously monitoring orientation using vestibular information. Neuropsychologia 40:373– 383; 2002.