Functional, independent ambulation places demands not only on the sensory-motor systems that control our balance and gait but also on cognitive systems.1 In everyday life, walking is often done in conjunction with other activities that place additional demands on cognitive processes, such as carrying a cup of hot coffee, remembering the items to get at the store, or talking on a cell phone. When doing 2 things at once, that is, dual task conditions, the individual may compromise performance in completing the tasks, compared to performance when each task is done by itself. The need to apply dual task conditions in physical therapy examinations has been increasingly realized, with the aim of preparing clients to participate in natural environments.2,3
Dual task deficits for functional motor skills performed while completing cognitive tasks have been documented in older adults and older adults with stroke.1,3–10 Cognitive tasks interfere with older adults' standing balance and walking.3 For those with stroke, sitting balance4 and standing balance5,6 can be compromised when paired with tasks that have a cognitive demand. Decrements in dual task performance for those with stroke have been reported for overground walking 6–10 and walking on a treadmill.1
The cognitive task selected in dual task paradigms varies from study to study. In some cases, tasks were selected that are known to assess a specific construct in cognition such as working memory.11 In the current study, we were interested in the everyday task of conversational speech. In addition to the motor demands of talking, speech generation requires attention and working memory.8 Using a methodology to examine age-related differences in speaking,12–15 we were interested in the rate of speech and specific components of speech including fluency, grammatical complexity, and semantic content.
Fluency includes the ability to select words, form sentences, and articulate.16,17 Diminished fluency is thought to relate to attention or memory deficits, or problems with articulation.16,17 Grammatical complexity describes the use of embedded and subordinate clauses, and it is correlated with working memory abilities.15 Semantic content examines the number of ideas and repetitions within spoken language. Deficits in delivery of content may relate to inefficient processing or limitations in working memory.15 Changes in the rate of speech or alterations in the components of speech may not be obvious to the casual listener. At some point, however, deterioration of speech can be noticeable, leading to frustration and confusion by the speaker and the listener.
Subjects may use different strategies to meet the demands of performing 2 tasks simultaneously. An individual could compromise performance on the motor tasks,11,18 the cognitive task,19,20 or both.21 Another strategy would be, if possible, to switch between the 2 tasks, rather than performing both simultaneously.21 Older adults compromised cognitive performance while maintaining safe ambulation when walking a narrow path19 and when controlling standing posture.20 Others have reported that older adults compromise performance in motor tasks while maintaining cognitive performance11 or even while demonstrating better cognitive performance.18 Adults with stroke compromised performance on both tasks when walking and performing a memory task6 or when walking and speaking.8,22 In contrast, others have reported that adults with stroke altered walking but not cognitive performance on a memory task.8
It is not known how subjects “decide” which task will be compromised when faced with the demands of doing 2 things at once. It has been suggested that cognitive performance will be maintained and postural stability sacrificed unless a threat of injury or limit of stability is reached.18,23 The selection of a strategy to accomplish 2 things at once safely demonstrates flexibility in control, a quality that may be preserved in healthy older adults.24
The purpose of this study was to examine how older adults with and without stroke meet the demands of walking while conversationally speaking. We hypothesized that participants would demonstrate dual task deficits in both walking and talking. We expected those with stroke to have greater dual task deficits in measures of walking and talking than older adults without stroke.
We used a quasi-experimental, within subject design comparing performance in single and dual task conditions of older adults without stroke and older adults with stroke. Each participant performed both single and dual task conditions, that is, speaking, walking, and speaking while walking. The data presented are part of a larger study examining dual task performance in older adults with and without stroke-related brain damage.
A total of 36 older adults were recruited for the study, 12 without stroke and 24 with stroke. To be included, participants had to be at least 50 years old, right-hand dominant,25 living in the community, and able to ambulate at least 7.6 m without the assistance of another person, with or without the use of a cane. They had to be free of any neurological disorder except for stroke for those in the group with stroke. Participants with stroke had to be at least 6 months post-onset. Those with stroke were excluded if they were aphasic, as determined by the Lexical Retrieval Tests of the Aphasia Diagnostic Profile.26
Participants with stroke were recruited from a university stroke registry. Those without stroke were recruited from the spouses of those with stroke and from a university registry of older adults willing to participate in research studies. Structural brain scans using magnetic resonance imaging (MRI) were obtained within 2 weeks of testing from 17 of the participants with stroke to confirm stroke-related pathology. If participants were unwilling or unable to undergo MRI, the diagnosis of stroke was determined from information from the registry or clinical presentation.
The study was approved by the university's institutional review board, and all participants provided oral and written consent. Data collection was done by a trained research assistant. Demographic data was collected including their age, sex, and years of education (Table 1). Tests were performed to determine the individual's ability to participate and to describe the sample, including a 10-m gait velocity test, and for those with stroke, the Stroke Impact Scale.27 A battery of cognitive tests was given including the Mini-Mental State Examination (MMSE)28 and the Shipley Vocabulary Test,29 which is a measure of verbal ability. Participants also were tested on components of the Wechsler Adult Intelligence Scales—Revised including Digits Forward, a measure of short-term memory; Digits Backwards, a measure of working memory; and Digit Symbol, a measure of psychomotor speed and visuospatial ability (Table 1).30 Participants performed walking and speaking in single and then dual task conditions. Rests, approximately 1- to 3-minute long, were provided between conditions to avoid fatigue. For those with stroke who were able to undergo MRI, the scans were scheduled on a separate day.
The walking and talking tasks were based on our previous work.22 Walking was performed on a 2-ft wide, irregular-shaped, elliptical pathway that was, at its largest extent, approximately 5 m by 3.5 m in diameter. The pathway was marked on an industrial carpet floor with brightly colored tape. Participants were instructed to walk at a comfortable pace, to avoid stepping on or over the tape, and to continue walking until told to stop. During walking and talking trials, they were not given any instructions to prioritize one activity or the other. They could choose to walk in a clockwise or counter-clockwise direction. No practice was provided. As the subject began walking, it was digitally videotaped and recorded for approximately 2 minutes for analysis. To avoid changes in walking as the participant slowed down, the participant was instructed to stop walking after the recording stopped.
Language samples were obtained using the methodology from previous studies of speech in older adults12–15 and adults with stroke.8,22 Speech was elicited with a variety of questions requiring participants to generate a minimum of a 2-minute response. One of the 12 questions was read asking the participant about their experiences or opinions, such as describing a favorite vacation, an individual they admire, or the most significant invention of the 20th century. The questions were randomized for each participant with the stipulation that no question was repeated for an individual. If participants paused or stopped talking during data collection, standardized verbal prompts were provided, for example, “Can you tell me more about this?” The baseline language sample was obtained while the participant was seated. For walking and talking trials, participants were given the question and then the walking instructions. Speech was digitally audio-recorded for approximately 2 minutes and time-locked with the videotaped walking data for later analysis.
For both walking and talking data, processing started 30 seconds after the subject initiated the task. For walking, the Noldus Video Observer (Noldus, 1997) was used to identify each foot contact from the videotapes. Foot contacts were identified, coded, and then time-locked to the audio recording. The steps were counted, and the rate was calculated as the cadence or number of steps per minute. Two research assistants were trained in this process. To establish reliability, 20% of the video recordings were analyzed independently by each research assistant. They were required to agree within 10 milliseconds on the onset and offset of all steps. Agreement for the number of steps and cadence was high, r > 0.98.
During walking and talking, codes were inserted that marked the onset of talking and pauses in talking. The percentage of time each participant was talking while walking was computed as a measure of the time when participants were actually performing both activities simultaneously.
Speech samples were analyzed using previously published methods.8,12–15,22 Audio tapes were transcribed, with nonlexical fillers, such as “uh,” “umm,” and “duh” excluded from the transcript. Lexical fillers, such as “you know” and “well” were retained. The words were counted, and the rate was calculated as the number of words per minute. Speech samples were coded into utterances, that is, a discernible pause in the subject's speech. Two additional research assistants were trained in this process. To establish reliability, 20% of the audio recordings were analyzed independently by each research assistant. For all language measures, an agreement was established, r > 0.90. Three areas of language were evaluated: fluency, grammatical complexity, and semantic content.
Three measures of fluency were calculated: (1) mean length of an utterance, (2) percentage of utterances that are grammatical sentences, and (3) percentage of utterances without lexical fillers.
Two measures of grammatical complexity were calculated: (1) mean clauses per utterance obtained by identifying each main and subordinate clause in each utterance and (2) developmental level that analyzes the inclusion of modifying clauses. A left-to-right processing model for speaking was used, such that clauses modifying the subject require more processing than clauses modifying the predicate.
Two measures of semantic content were calculated: (1) propositional density, which is the average number of propositions per 100 words and (2) type-token ratio, which is the ratio of the number of different words in the sample to the total number of words in the sample.
Analysis was completed using SPSS software, version 16.0 for Windows (SPSS Inc, Chicago, Illinois). Means and standard deviations for all variables were calculated for groups in both single and dual task conditions. To determine whether there was a dual-task effect for older adults without stroke and older adults with stroke, cadence, speech rate, and each language measure were analyzed for each group using a paired t test to examine the effects of condition (single task, dual task).
Group differences (older adults without stroke, older adults with stroke) for the percentage of time participants were performing both activities at the same time and dual task costs (DTC) for cadence, speech rate, and each language measure were analyzed with independent sample t tests. The DTC controls for baseline performance differences and was calculated using the following formula:
A negative DTC indicates performance was better in the single task condition. A positive DTC indicates performance was better in the dual task condition. We present means with standard deviations in parentheses, unless otherwise indicated.
Demographic data are provided in Table 1 for each group of participants. Four of the adults with stroke used a cane for walking. They used the cane for both conditions, that is, walking and walking while talking. Each of these individuals used a cane for their daily walking activities and had done so for 9 to 36 months since their stroke.
The results of analysis for the effect of the dual task condition on rates of walking (cadence) and on talking (speech rate) are provided in Table 2. For both groups, dual task effects were found for cadence but not for speech rate.
The effect of the dual task condition on language components is presented in Table 3. More than 65% of the subjects had 0% of utterances without lexical fillers in single and dual task conditions, so this measure of fluency was dropped from further analysis. For the participants without stroke, there were no differences between single and dual task performance for any language measure. Those with stroke had poorer performance in the dual task condition compared with the single task condition in both measures of grammatical complexity, that is, mean clauses per utterance and developmental level and in both measures of semantic content, that is, propositional density and type-token ratio.
Effect of Group for Simultaneous Performance and Dual Task Costs
The effect of group on simultaneous performance and DTC were as follows: Participants without stroke were able to walk and talk simultaneously 96.3% (2.2%) of the time. Those with stroke only did both tasks at the same time for 90.6% (8.1%) of the time. This resulted in a significant group difference, P = .02.
There were no group differences for DTC for cadence or speech rate (Table 4). For the language measures, group differences were restricted to the 2 measures of grammatical complexity, with those with stroke showing greater DTC than those without stroke (Table 4).
Individuals may produce different strategies to meet the demands of performing 2 tasks simultaneously. Performance on one or both tasks may be compromised by reducing the speed or accuracy of the task. Another strategy could be to stop doing both tasks simultaneously, instead switching performance between one task and the other. Contrary to our hypothesis, the adults without stroke only altered the speed of walking, without a change in the rate or pattern of their speech. There was a different response to the demands of walking and talking for those with stroke. These subjects slowed their walking and altered their speech, specifically by diminishing the grammatical complexity and semantic content. Furthermore, those with stroke spent less time in walking and talking simultaneously.
Many studies using dual task paradigms do not report time spent on both tasks concurrently. In examining this measure, we demonstrate that 1 strategy to deal with performing 2 tasks at the same time may be to actually perform only 1 task at a time. This is the basis for the clinical test for frail elderly in the “stops walking when talking” test.31 Examination of performance on concurrent performance and performance of each task is revealing. Some reports of dual task deficits have measured only the motor task without reporting data on performance on the cognitive task.32–34 When both tasks have been analyzed, some have shown decrements in both tasks,35,36 only the cognitive task,37 or only the motor task.11,23,37
Walking a straight pathway, adults aged 65 to 69 years have a cadence of approximately 115.38 Given the constraints of the walking path in the current study, cadence was only 98.9 and 82.6 without talking for those without stroke and with stroke, respectively. Both groups of subjects reduced their cadence even more in response to the demands of the dual task condition. Cadence is generally linearly related to gait speed; decreasing cadence contributing to a decrease in gait velocity.39 Some have argued that slowing gait is a positive strategy to gain stability.40 Walking slower than the individual's preferred speed can increase stability even though variability can increase,41,42 although some have shown that walking slower than the preferred speed is not more stable.43–45 We did not measure stability in the current study, but did examine stumbles from the video recordings. Only 2 participants with stroke stumbled, and each did this on only 1 step when walking without talking. Both regained their balance independently. There were no other obvious incidents of loss of balance, suggesting that decreasing walking speed was a safe resolution to meet the demands of walking and talking. This adaptation may be similar to that reported in studies of standing balance during cognitive tasks. Older adults will allow an increase in postural instability to successfully complete a cognitive task, until the postural instability breaches some perceived level of safety.20,24,35
Only those with stroke altered the pattern of their speech, changing the grammatical complexity and semantic content. These changes are subtle and would not be obvious in the clinical setting. However, conversational speech may suffer to a greater degree in more demanding motor activities. Also, our subjects with stroke were highly recovered and living in the community. More profound changes in speech may occur if individuals have greater sensory-motor impairment than those in the current study. Of note, the changes observed in grammatical complexity and fluency may relate to deficits in working memory. Examination of the screening tests revealed that there were no group differences in the Digit Span Backward test, a measure of working memory. It remains to be seen whether these specific patterns of altered speech relate to a specific cognitive skill.
We selected tasks that are ecologically meaningful. Walking straight forward, a short distance is a functional task that has been used for clinical testing of dual task abilities.3 We suggest that walking is not always conducted in a free, unrestricted environment. Walkways may be cluttered, narrow, and contain numerous turns. The irregular-shaped, narrow path in the current study was designed to mimic these conditions. Dual task deficits may be uncovered in these more challenging walking tasks. Conversational speech during walking is a common, well-practiced task. Arguably, independent functional ambulation would include the ability to walk and talk at the same time.
There are limitations to our study. Subjects without stroke had a mean age of 72.7 years and those with stroke had a mean age of 66.5 years, representing a relatively young group. We did not obtain MRI brain scans for participants who had no history of stroke and for 7 of our 24 subjects with stroke. We cannot rule out the presence of brain damage unrelated to stroke or silent infarcts; the presence of brain infarcts have been reported in 10% of community dwelling adults who have no clinical signs of stroke.46 There were statistically significant group differences in 10-m gait velocity, education level, and 3 cognitive tests, that is, MMSE, digit symbol, and Shipley vocabulary. We do not know whether adults with stroke who score higher in these tests would show the same difficulties in walking and talking as the cohort in the present study. Multiple t tests increase the chance of a type 1 error. A conservative analysis using a Bonferroni adjustment for the walking and talking data would result in an alpha level of .002; this should be considered in interpretation of the data. The protocol and data processing for our language measures were time-consuming and are not amenable to clinical practice. It is possible that with more demanding tasks or for those with greater motor or speech deficits, changes in speech would be evident to the clinician. Dual task deficits with walking are associated with the incidence of falls in older adults.47 A similar relationship for those with stroke has been reported.6 We did not examine fall history in our subjects, but investigations of potential dual task deficits in these populations are critical, particularly with common functional activities that are performed simultaneously.
Daily life includes many situations where 2 tasks are performed at the same time, and walking and talking are among the most common. Older adults with and without stroke may be challenged by the demands of talking while walking a narrow, irregular pathway and slow down their cadence. This accommodation may be a reasonable strategy to ensure safety. Those with stroke also may make subtle changes in speech. Clinicians should consider the impact of conversational speech during standardized testing that is not designed to assess dual task performance. However, additional testing that includes dual task activities should be considered, with attention to combining activities that the client performs together in daily life. In addition to revealing performance deficits, the clinician should consider the impact of the strategy selected to safely meet the demands of performing 2 tasks at once. There is evidence that dual task performance can be improved in older adults with practice, both without35 and with stroke,10 indicating that interventions targeted at dual task experiences can diminish functional deficits when performing 2 activities at the same time.
This study was supported by a Grant-in-Aid from the Heartland Affiliate of the American Heart Association to Patricia S. Pohl.
1. Regnaux JP, David D, Daniel O, Ben Smail D, Combeaud M, Bussel B, Evidence for cognitive processes involved in the control of steady state of walking in healthy subjects and after cerebral damage. Neurorehabil Neural Repair, 2005;19:125–132.
2. McCulloch K, Attention and dual-task conditions: physical therapy implications for individuals with acquired brain injury. J Neurol Phys Ther. 2007;31:104–118.
3. McCulloch KL, Mercer V, Guiliani C, Marshall S, Development of a clinical measure of dual-task performance in walking: reliability and preliminary validity of the walking and remembering test. J Geriatric Phys Ther, 2009;32:2–9.
4. Harley C, Boyd JE, Cockburn J, et al., Disruption of sitting balance after stroke: influence of spoken output. J Neurol Neurosurg Psychiatry. 2006;77:674–676.
5. Bensoussan L, Viton JM, Schieppati M, et al., Changes in postural control in hemiplegic patients after stroke performing a dual task. Arch Phys Med Rehabil. 2007;88:1009–1015.
6. Hyndman D, Ashburn A, Yardley L, et al., Interference between balance, gait and cognitive task performance among people with stroke living in the community. Disabil Rehabil. 2006;28:849–856.
7. Cockburn J, Haggard P, Cock J, et al., Changing patterns of cognitive-motor interference (CMI) over time during recovery from stroke. Clin Rehabil. 2003;17:167–173.
8. Plummer-D'Amato P, Altmann LJP, Saracino D, et al., Interactions between cognitive tasks and gait after stroke: a dual task study. Gait Posture. 2008;27:683–688.
9. Yang Y-R, Chen Y-C, Lee C-S, et al., Dual-task-related gait changes in individuals with stroke. Gait Posture. 2007;25:185–190.
10. Yang Y-R, Wang R-Y, Chen Y-C, et al., Dual-task exercise improves walking ability in chronic stroke: a randomized controlled trial. Arch Phys Med Rehabil. 2007;88:1236–1240.
11. Lövdén M, Schaefer S, Pohlmeyer AE, Lindenberger U, Walking variability and working-memory load in aging: a dual-process account relating cognitive control to motor control performance. J Gerontol Psychol Sci. 2008;63:P121–P128.
12. Cheung H, Kemper S, Competing complexity metrics and adults' production of complex sentences. Appl Psycholinguistics. 1992;13:53–76.
13. Kemper S, Herman RE, Lian CHT, The costs of doing two things at once for young and older adults: talking while walking, finger tapping, and ignoring speech or noise. Psychol Aging, 2003;18:181–192.
14. Kemper S, Kynette D, Rash S, Sprott R, O'Brien K, Life-span changes to adults' language: effects of memory and genre. Appl Psycholinguistics. 1989;10:49–66.
15. Kemper S, Sumner A, The structure of verbal abilities in young and older adults. Psychol Aging. 2001;16:312–322.
16. Levelt WJM, Accessing words in speech production: stages, processes, and representations. In: WJM Levelt, ed. Lexical Access in Speech Production. Oxford, UK: Blackwell; 1992:1–22.
17. Maclay H, Osgood CE, Hesitation phenomena in spontaneous English speech. Word. 1959;15:19–44.
18. Verrel J, Lövdén M, Schellenbach M, Schaefer S, Lindenberger U, Interacting effects of cognitive load and adult age on the regularity of whole-body motion during treadmill walking. Psychol Aging. 2009;24:75–81.
19. Li KZH, Lindenberger U, Freund AM, Baltes PT, Walking while memorizing: a SOC study of age-related differences in compensatory behavior under dual-task conditions. Psychol Sci. 2001;12:230–237.
20. Rapp MA, Krampe RT, Baltes PB, Adaptive task prioritization in aging: selective resource allocation to postural control is preserved in Alzheimer disease. Am J Geriatr Psychiatry. 2006;14:52–61.
21. Siu KC, Lugade V, Chou LS, et al., Dual-task interference during obstacle clearance in healthy and balance-impaired older adults. Aging Clin Exp Res. 2008;20:349–354.
22. Kemper S, McDowd J, Pohl P, et al., Revealing language deficits following stroke: the cost of doing two things at once. Aging Neuropsychol Cogn. 2006;13:115–139.
23. Shumway-Cook A, Woollacott M, Kerns KA, 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. 1997;52A:M232–M240.
24. Doumas M, Smolders C, Krampe RT, Task prioritization in aging: effects of sensory information on concurrent posture and memory performance. Exp Brain Res. 2008;187:275–281.
25. Bryden M, Laterality: Functional Asymmetry in the Intact Brain. New York, NY: Academic Press; 1982.
26. Helm-Estabrooks N, Aphasia Diagnostic Profiles. Austin, TX: Pro-Ed Inc; 1992.
27. Duncan PW, Wallace D, Lai SM, et al., The Stroke Impact Scale version 2.0: evaluation of reliability, validity, and sensitivity to change. Stroke. 1999;30:2131–2140.
28. Folstein MF, Folstein SE, McHugh PR, et al., Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatry Res. 1975;12:189–198.
29. Shipley WC, A self-administered scale for measuring intellectual impairment and deterioration. J Psychol. 1940;9:371–377.
30. Wechsler D. The Measurement and Appraisal of Adult Intelligence. Baltimore, MD: Williams & Willkins; 1958.
31. Lundin-Olsson L, Nyberg L, Gustafson Y, “Stops walking when talking” as a predictor of falls in elderly people. Lancet. 1997;349:617.
32. Brown LA, Shumway-Cook A, Woollacott MH, Attentional demands and postural recovery: the effects of aging. J Gerontol Med Sci. 1999;54A:M165–M171.
33. Kelly VE, Schrager MA, Price R, et al.. Age-associated effects of a concurrent cognitive task on gait speed and stability during narrow-based walking. J Gerontol Med Sci. 2008;63A:1329–1334.
34. Priest AW, Salamon KB, Hollman JH, Age-related differences in dual task walking: a cross sectional study. J Neurol Eng Rehabil. 2008;5:29.
35. Doumas M, Rapp MA, Krampe RT, Working memory and postural control: adult age differences in potential for improvement, task priority, and dual tasking. J Gerontol Psychol Sci. 2009;64B:193–201.
36. Srygley JM, Mirelman A, Herman T, et al., When does walking alter thinking? Age and task associated findings. Brain Res. 2009;1253:92–99.
37. Dennis A, Dawes H, Elsworth C, et al., Fast walking under cognitive-motor interference conditions in chronic stroke [published online ahead of print June 13, 2009]. Brain Res. 2009;1287:104–110.
38. Lord SR, Lloyd DG, Li SK, Sensori-motor function, gait patterns and falls in community-dwelling women. Age Ageing. 1996;25:292–299.
39. Perry J, Gait Analysis: Normal and Pathological Function. Thorofare, NJ: Slack; 1992:432.
40. Rogers HL, Cromwell RL, Grady JL, Adaptive changes in gait of older and younger adults as responses to challenges to dynamic balance. J Aging Phys Act. 2008;16:85–96.
41. England SA, Granata KP, The influence of gait speed on local dynamic stability of walking. Gait Posture. 2007;25:172–178.
42. Kang HG, Dingwell JB, Effects of walking speed, strength, and range of motion on gait stability in healthy older adults. J Biomech. 2008;41:2899–2905.
43. Jordan K, Challis JH, Newell KM, Walking speed influences on gait cycle variability. Gait Posture. 2007;26:128–134.
44. Latt MD, Menz HY, Fung VS, Lord SR, Walking speed, cadence and step length are selected to optimize the stability of head and pelvis accelerations. Exp Brain Res. 2008;184:201–209.
45. Bruijn SM, van Dieën JH, Meijer OG, Beek PJ, Is slow walking more stable? J Biomech. 2009;42:1506–1512.
46. Das RR, Seshadri S, Beiser AS, et al., Prevalence and correlates of silent cerebral infarcts in the Framingham offspring study. Stroke. 2008;39:2929–2935.
47. Beauchet O, Annweiler C, Allali G, et al.. Recurrent falls and dual-task related decrease in walking speed: Is there a relationship? J Amer Geriatric Soc. 2008;56:1265–1269.