- Email: [email protected]
Cutting a long story short: Reaction times in acute stroke are associated with longer term cognitive outcomes
Journal of the Neurological Sciences 322 (2012) 102–106
Contents lists available at SciVerse ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Cutting a long story short: Reaction times in acute stroke are associated with longer term cognitive outcomes Toby B. Cumming a, b,⁎, Amy Brodtmann a, b, David Darby b, c, Julie Bernhardt a a b c
Stroke Division, Florey Neuroscience Institutes, Melbourne, Australia Cognitive Neuroscience Division, Florey Neuroscience Institutes, Melbourne, Australia Department of Medicine, University of Melbourne, Australia
a r t i c l e
i n f o
Article history: Received 31 January 2012 Received in revised form 4 June 2012 Accepted 2 July 2012 Available online 20 July 2012 Keywords: Cerebrovascular disease Cognitive impairment Attention Psychomotor Neuropsychology
a b s t r a c t Background: The viability and usefulness of cognitive assessment in acute stroke have been questioned, with practical challenges arising from the focal nature of neurological deficits as well as heterogeneity in arousal state. We aimed to test whether acute measures of attention correlate with attentional function at 3 months post-stroke. Methods: Patients with confirmed stroke completed 2 computerised cognitive tasks (CogState) within 2 weeks of stroke. The tasks were a simple reaction time task (Detection) and a choice reaction time task (Identification) that required a button press to visual stimuli (playing cards). Each task took approximately 4 min. The Montreal Cognitive Assessment (MoCA) and an extended neuropsychological battery were administered at 3 months post-stroke. Results: Thirty-three patients (mean age 75.5 years, SD 11.9) participated in this preliminary study. Correlations indicated that both Detection speed (r=−0.73, p b 0.001) and Identification speed (r=−0.61, p =0.007) at baseline were associated with attentional function at 3 months, as measured by established neuropsychological tests (Trails-A, Digit span, Digit symbol). In addition, Detection speed at baseline was correlated with total 3-month MoCA score (r=−0.54, p=0.012). Conclusion: Simple and brief computerised assessment of attentional function in acute stroke is possible and is related to longer term attentional and cognitive performance. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Approximately two thirds of stroke survivors experience cognitive impairment [1], with many going on to develop dementia [2]. Stroke appears to have particularly strong effects in certain cognitive domains, with impairments being most pronounced in the areas of attention/executive function and speed of processing [3]. As cognitive domains are not separate but inter-dependent, deficits in one area can influence performance in other areas. For example, the ability to direct, shift and sustain attention underpins many mental skills, including aspects of memory [4]. Thus it may be possible to cut the long story of multi-domain neuropsychological testing short. While the broad classification of ‘cognitive impairment’ after stroke has been associated with increased death and disability [5], greater rates of institutionalisation [6] and poorer rehabilitation outcomes [7], processing speed alone appears related to functional outcome after stroke [8]. Early identification of post-stroke cognitive deficits has the potential to guide therapy (e.g., ‘attention process training’ for attentional deficits ⁎ Corresponding author at: Melbourne Brain Centre, 245 Burgundy St, Heidelberg, 3084, Australia. Tel.: +61 390357152; fax: +61 394962251. E-mail address: [email protected] (T.B. Cumming). 0022-510X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2012.07.004
[9]) and inform rehabilitation strategies. There are practical barriers, however, to assessing cognition early after stroke. Many acute stroke patients have reduced level of consciousness, dysphasia, visual deficits and weakness in the hand and arm, all of which can limit their ability to complete cognitive assessments. In one study, researchers attempted to demonstrate the feasibility of neuropsychological testing between 4 and 20 days after first ischaemic stroke. Yet of the 73 included patients, only 57 were examined, and only 44 could complete 82% of the tests [10]. Shorter screening tools such as the Mini-Mental State Examination and the Montreal Cognitive Assessment have been employed in acute stroke [11,12], although their reliance on written and verbal responses means many moderate and severe stroke patients cannot complete them. There is also some doubt about whether cognitive status in the often medically unstable acute stage accurately reflects cognitive or functional status in the longer term. Early cognitive deficits have been related to longer term cognitive impairment, dependence in activities of daily living, depression and lower quality of life [13,14], but more evidence is required. It is possible that simple computerised cognitive tasks can overcome some of the difficulties in the assessment of acute stroke patients. Computerised reaction time tasks have been used in stroke samples to test attention, processing speed and working memory [15,16], but
T.B. Cumming et al. / Journal of the Neurological Sciences 322 (2012) 102–106
103
2. Methods
flipped over. The card was either a red or a black joker. The task ended after 30 correct trials were completed. Visual and auditory feedback differentiated correct and incorrect responses, and anticipatory responses (button press before card flipped over) triggered a distinctive error sound. Accuracy and reaction times for correct responses were recorded. The MoCA [22] is scored out of 30 and includes items on visuospatial ability, executive function, naming, memory, attention, language, abstract thinking and orientation. It has been recommended as a useful screen for vascular cognitive impairment [23], and appears to be slightly more sensitive than the Mini-Mental State Examination after stroke [12]. The attentional tasks administered as part of the 3-month neuropsychological battery were Trail-Making part A (draw a line between numbers 1 to 25 in ascending order), Digit Span (repeat number strings) and Digit Symbol (match symbols to numbers according to a specified code). Z-scores were derived for each patient on each of these 3 tasks using well-established age-based population norms. These 3 z-scores were then averaged to form a single summed z-score for attentional function.
2.1. Participants
2.4. Data analysis
Patients who were admitted to the acute stroke unit at the Austin Hospital with completed stroke (ischaemic or intracerebral haemorrhage) were eligible for this study. Patients were excluded if they: (a) were younger than 18 years old, (b) suffered the stroke more than 2 weeks previously, (c) required an interpreter, or (d) had major visual, hearing or receptive language impairments which prevented completion of the cognitive tasks. All patients provided informed consent prior to inclusion in the study. Relevant demographic and clinical characteristics, including age, stroke severity using the NIHSS [20], Oxfordshire classification [21] and lesion side, were extracted from the patient's medical record. Ethical approval for this study was obtained from the Austin Health Human Research Ethics Committee.
For the computerised tasks, a base 10 logarithmic transformation was applied to normalise the distributions of mean reaction time. If a patient made an error on 50% or more of the trials, their reaction time data for that task were excluded from analysis. Pearson correlations were computed between acute reaction times and a composite 3-month attentional score. Multivariate linear regressions were then conducted that included age and baseline stroke severity (NIHSS) as confounding variables. Additional correlations were performed to test the relationship between baseline stroke severity (NIHSS) and the acute reaction time data. As there were no previous data on which to base an estimate of effect size, no formal sample size calculation was performed. We aimed to include 30 patients to justify including 3 independent variables in regression, so attempted to recruit 33 to allow for a 10% dropout rate.
typically not in the acute stage. The CogState computerised battery (CogState Ltd, Melbourne, Australia) is composed of tasks that were derived from the cognitive paradigms of psychomotor function, attention and working memory. These reaction time tasks are short, straightforward and not reliant on expressive language; they require button press responses to the presentation of playing cards on a computer screen. Research in other populations has shown the CogState tasks to be valid and reliable [17], free of practice effects in serial testing [18] and sensitive to change [19]. In this preliminary study, we hypothesised that response times on simple and choice reaction time tasks in the acute stage (within 2 weeks) of stroke would be correlated with standard measures of attention at 3 months post-stroke. A second hypothesis was that acute reaction times would be correlated with total score on the Montreal Cognitive Assessment (MoCA) at 3 months.
2.2. Procedure 3. Results The baseline testing session took place on the acute stroke ward immediately following written informed consent. The cognitive tasks were presented on a laptop computer that had 2 large (approximately 6 cm in diameter) external response keys (the ‘yes’ and ‘no’ buttons) attached. Each patient was either sitting up in bed or sitting in a chair next to their bed for the testing (wherever they were originally placed). Curtains were drawn to minimise interruptions and to encourage the patient to focus on the task. At 3 months post-stroke, 2 follow-up visits were made to the patient's residence. During the first visit, a cognitive screening tool (MoCA) was completed. At the second visit, a comprehensive neuropsychological battery was administered that included attentional tasks. The same researcher (TBC) administered the neuropsychological battery and the acute testing session, but he remained blind to the results of the acute cognitive tasks until the end of the study. 2.3. Outcome measures The computerised tasks administered in the acute stage were the Detection task and the Identification task from the CogState battery. Both tasks were preceded by instructions read aloud by the experimenter and a series of practice items for task familiarisation. A playing card was presented face down in the centre of the screen and, after an interval that varied randomly between 2.5 and 3.5 s, instantly turned face up. The Detection task is a simple reaction time test of psychomotor function and speed of processing; the person was instructed to press the ‘yes’ button as soon as possible after the card on screen flipped over. The card was always the same joker card. The task ended after 35 correct trials were completed. The Identification task is a choice reaction time test of visual attention; the person was instructed to press the ‘yes’ button if the card is red and the ‘no’ button if the card is not red as soon as possible after it
Thirty-three patients participated; demographics and stroke characteristics are shown in Table 1. The predominantly male sample had a mean age of 75.5 and a mean acute NIHSS score of 5.1. 3.1. Feasibility Baseline testing was completed at a mean of 5.4 days post-stroke (SD =2.9), with a range between 1 and 12 days. Each of the computerised tasks can be completed within 3 min, including instructions and practice items. A number of stroke patients took longer to understand the instructions, made a number of errors and had slowed responses, so these tasks often took approximately 4 min each. For the Detection task, 32 patients had complete data. One patient finished the task, but their data could not be retrieved due to equipment failure. Six patients did not get more than 50% correct and were therefore excluded from calculations, leaving data from 26. For the Identification task, 31 patients had complete data. One patient finished the task, but their data could not be retrieved due to equipment failure. Another patient had to stop testing after the Detection task due to dizziness and weakness, and thus did not complete the Identification task. Seven patients failed to perform above chance (50%) and were therefore excluded from calculations, leaving data from 24. Log10 reaction times are presented in Table 2. Hypothesis 1. Detection speed at baseline was significantly correlated with attentional function z-score at 3 months (n =21; r=−0.73, pb 0.001). This relationship is illustrated in Fig. 1. In multivariable regression, Detection speed at baseline was significantly associated with attentional function at 3 months (t=−5.24, pb 0.001), with age (p=0.15) and stroke severity (p=0.07) included as independent variables. This
104
T.B. Cumming et al. / Journal of the Neurological Sciences 322 (2012) 102–106
Table 1 Patient characteristics (N = 33). N (%)a Age — mean (SD), range Male NIHSS score — mean (SD), range Mild (1–7) Moderate (8–15) Oxfordshire stroke classification TACI PACI POCI LACI ICH Left-sided lesion Right-sided lesion
75.5 (11.9), 45–95 25 (76) 5.1 (3.2), 1–15 26 (79) 7 (21) 1 (3) 18 (55) 9 (27) 3 (9) 2 (6) 16 (52) 15 (48)
NIHSS — National Institutes of Health Stroke Scale; TACI — total anterior circulation infarct; PACI — partial anterior circulation infarct; POCI – posterior circulation infarct; LACI — lacunar infarct; ICH — intracerebral haemorrhage. a N (%) unless otherwise specified. Lesion side missing for 2 patients. Fig. 1. Relationship between reaction time on the Detection (DET) task at baseline and summed 3-month attention z-score. 2
model was highly significant (R =0.66; F=10.8, pb 0.001). Identification speed at baseline was also correlated with attentional function at 3 months (n=18; r=−0.61, p=0.007). In regression, Identification speed at baseline was significantly related to 3-month attentional function (t = − 3.50, p = 0.004), with age (p = 0.12) and stroke severity (p = 0.59) included. This regression model was also significant (R 2 = 0.51; F = 4.8, p = 0.017). Hypothesis 2. Detection speed at baseline was significantly correlated with MoCA score at 3 months (n = 21; r = − 0.54, p = 0.012). Fig. 2 shows this relationship. In multivariable regression, baseline Detection speed was significantly associated with MoCA score at 3 months (t = − 2.39, p = 0.029), with age (p = 0.11) and stroke severity (p = 0.61) included. This regression model was significant (R 2 = 0.40; F = 3.7, p = 0.031). Identification speed at baseline was moderately correlated with MoCA score at 3 months, but this did not reach significance (n = 18; r = − 0.39, p = 0.11). In multivariable regression, Identification speed at baseline was not significantly related to MoCA at 3 months (t=−1.27, p=0.23), with age (p=0.14) and stroke severity (p=0.75) included. This regression model was not significant (R2 =0.30; F=2.0, p=0.16). There was no significant correlation between acute stroke severity (NIHSS) and Detection speed (n= 26; r = 0.04, p = 0.84; see Fig. 3) or Identification speed (n= 24; r = 0.18, p = 0.41) at baseline.
the first 3 weeks after stroke and longer term cognitive impairment. This prior study, however, employed standard neuropsychological testing in both the acute setting and long-term follow-up. Our findings indicate that simple and choice reaction time tasks administered in the first week following stroke can also be prognostic for longer term cognitive performance. One possible explanation is that acute reaction times were merely a proxy for stroke severity. This is unlikely, however, given the lack of correlation between acute NIHSS score and speed on the Detection and Identification tasks. We were surprised to find no relationship here, as stroke severity is normally a strong predictor of not only cognitive but other functional outcomes. It appears that the computerised cognitive tasks provide important information beyond a clinical impression of neurological impairment. Score on the NIHSS scale, which incorporates a wide range of neurological deficits including level of consciousness, dysphasia, neglect, visual field loss, motor weakness, ataxia, dysarthria and sensory loss, does not necessarily account for differences in attention and speed of processing. There is growing awareness of the role that stroke and vascular risk factors play in the development of cognitive impairment and dementia [2,24]. Ideally, there would be cognitive assessments that could be employed soon after stroke (to obtain a post-stroke baseline) and that could be repeated multiple times subsequently (to track recovery or decline). The Detection and Identification tasks used here have these
4. Discussion We found that the reaction times of acute stroke patients on simple computerised tasks were related to attentional function assessed at 3 months post-stroke. These correlations could not be explained by age or stroke severity, as the associations remained when these variables were accounted for in multivariable regression. We also documented a significant relationship between speed on the Detection task (simple reaction time) at baseline and total score on the MoCA (a global cognitive screening tool) at 3 months post-stroke. There is some precedent for these results, with Nys et al. [13] reporting a strong relationship between domain-specific cognitive abilities tested in
Table 2 Log10 reaction times (and millisecond equivalents) for the Detection (DET) and Identification (IDN) tasks at baseline.
DET IDN
N
Mean log10 RT
SD
Min
Max
Mean ms
26 24
2.64 2.83
0.17 0.12
2.37 2.54
2.98 3.05
437 679
Fig. 2. Relationship between reaction time on the Detection (DET) task at baseline and total MoCA score at 3 months.
T.B. Cumming et al. / Journal of the Neurological Sciences 322 (2012) 102–106
105
Acknowledgements We thank all participants for their time and effort. We also thank Debbie Hansen and Karen Moss for helping collect the data. This work was supported by research grants from the National Stroke Foundation and from the Equity Trustees Preston & Loui Geduld Trust Fund. Dr Cumming is funded by a National Heart Foundation Postdoctoral Research Fellowship. All research undertaken at the Florey Neuroscience Institutes is supported by infrastructure funding from the Department of Innovation (DOBI).
References
Fig. 3. Relationship between reaction time on the Detection (DET) task at baseline and stroke severity (NIHSS score) at baseline.
desirable characteristics. As straightforward reaction time tasks, they do not cover the range of cognitive domains that a neuropsychological battery covers. They may be very useful, though, as measures of attentional function and possibly therefore as surrogate markers for general cognition. The relationship between lesion location and profile of cognitive deficits is not always predictable, and impairment may arise separate to any focal injury directly attributable to the stroke. It may be that infarction anywhere in the brain results in disruption to the widely distributed attention network – or to its critical and abundant neurotransmitters – which is then reflected in slower processing, more diffuse activation and more disorganised thinking. Simple and choice reaction times might be markers for the extent of this attention network disruption. As a group, the acute stroke patients tested here exhibited slower reaction times than a group of healthy older people tested on the same Detection and Identification tasks. Fredrickson et al. [25] reported data for a group of 301 community-dwelling people with mean age of 62. The mean age in our stroke sample was 76. Although some slowing would be expected purely due to age, the patients in our study were markedly slower. With log10 reaction times translated to milliseconds, the stroke patients took longer than the community sample in Detection (mean 437 ms versus 331 ms) and Identification (mean 679 ms versus 525 ms). Several limitations to the current study should be noted. The sample was small, and as such our results require replication in larger groups. The sample was made up of mild and moderate stroke patients only, and our findings therefore lack generalisability to more severe patients. While the CogState tasks appear more feasible than many other cognitive assessment options, there are still acute stroke patients whose level of consciousness or receptive dysphasia makes completing the tasks impossible. The patients were not sampled consecutively, although this should not pose a major problem as the analyses presented here are mostly within-subject. Finally, we have not examined the connection between lesion location and task performance, and therefore can make no conclusions about the influence of stroke topography. Our results indicate that simple and brief computerised assessment of attentional function in acute stroke is possible and is related to longer term attentional and cognitive performance. The exciting prospect that reaction time tasks can be used not only to track attentional performance over time following stroke but also as surrogate markers for more general cognitive function requires further evaluation in larger studies.
Conflict of interest Dr. Darby is a shareholder and consultant for CogState Ltd.
[1] Jin YP, Di Legge S, Ostbye T, Feightner JW, Hachinski V. The reciprocal risks of stroke and cognitive impairment in an elderly population. Alzheimers Dement 2006;2:171–8. [2] Pendlebury ST, Rothwell PM. Prevalence, incidence, and factors associated with pre-stroke and post-stroke dementia: a systematic review and meta-analysis. Lancet Neurol 2009 Nov;8(11):1006–18. [3] Feigin VL, Barker-Collo S, Parag V, Senior H, Lawes CM, Ratnasabapathy Y, et al. Auckland Stroke Outcomes Study. Part 1: gender, stroke types, ethnicity, and functional outcomes 5 years poststroke. Neurology 2010;75(18):1597–607. [4] McNab F, Klingberg T. Prefrontal cortex and basal ganglia control access to working memory. Nat Neurosci 2008;11:103–7. [5] Patel MD, Coshall C, Rudd AG, Wolfe CD. Cognitive impairment after stroke: clinical determinants and its associations with long-term stroke outcomes. J Am Geriatr Soc 2002;50(4):700–6. [6] Pasquini M, Leys D, Rousseaux M, Pasquier F, Henon H. Influence of cognitive impairment on the institutionalisation rate 3 years after a stroke. J Neurol Neurosurg Psychiatry 2007;78(1):56–9. [7] Heruti RJ, Lusky A, Dankner R, Ring H, Dolgopiat M, Barell V, et al. Rehabilitation outcome of elderly patients after a first stroke: effect of cognitive status at admission on the functional outcome. Arch Phys Med Rehabil 2002;83(6):742–9. [8] Barker-Collo S, Feigin VL, Parag V, Lawes CM, Senior H. Auckland Stroke Outcomes Study. Part 2: cognition and functional outcomes 5 years poststroke. Neurology 2010;75(18):1608–16. [9] Barker-Collo SL, Feigin VL, Lawes CM, Parag V, Senior H, Rodgers A. Reducing attention deficits after stroke using attention process training: a randomized controlled trial. Stroke 2009;40(10):3293–8. [10] van Zandvoort MJ, Kessels RP, Nys GM, de Haan EH, Kappelle LJ. Early neuropsychological evaluation in patients with ischaemic stroke provides valid information. Clin Neurol Neurosurg 2005;107(5):385–92. [11] Dong Y, Sharma VK, Chan BP, Venketasubramanian N, Teoh HL, Seet RC, et al. The Montreal Cognitive Assessment (MoCA) is superior to the Mini-Mental State Examination (MMSE) for the detection of vascular cognitive impairment after acute stroke. J Neurol Sci 2010;299(1–2):15–8. [12] Godefroy O, Fickl A, Roussel M, Auribault C, Bugnicourt JM, Lamy C, et al. Is the Montreal Cognitive Assessment Superior to the Mini-Mental State Examination to detect poststroke cognitive impairment?: a study with neuropsychological evaluation. Stroke 2011;42(6):1712–6 (2011 June 1). [13] Nys GM, van Zandvoort MJ, de Kort PL, van der Worp HB, Jansen BP, Algra A, et al. The prognostic value of domain-specific cognitive abilities in acute first-ever stroke. Neurology 2005;64(5):821–7. [14] Nys GM, van Zandvoort MJ, van der Worp HB, de Haan EH, de Kort PL, Jansen BP, et al. Early cognitive impairment predicts long-term depressive symptoms and quality of life after stroke. J Neurol Sci 2006;247(2):149–56. [15] Ballard C, Stephens S, Kenny R, Kalaria R, Tovee M, O'Brien J. Profile of neuropsychological deficits in older stroke survivors without dementia. Dement Geriatr Cogn Disord 2003;16(1):52–6. [16] Quaney BM, Boyd LA, McDowd JM, Zahner LH, He J, Mayo MS, et al. Aerobic exercise improves cognition and motor function poststroke. Neurorehabil Neural Repair 2009 Nov;23(9):879–85. [17] Collie A, Maruff P, Makdissi M, McCrory P, McStephen M, Darby D. CogSport: reliability and correlation with conventional cognitive tests used in postconcussion medical evaluations. Clin J Sport Med 2003;13(1):28–32. [18] Falleti MG, Maruff P, Collie A, Darby DG. Practice effects associated with the repeated assessment of cognitive function using the CogState battery at 10-minute, one week and one month test-retest intervals. J Clin Exp Neuropsychol Off J Int Neuropsychol Soc 2006;28(7):1095–112. [19] Silbert BS, Maruff P, Evered LA, Scott DA, Kalpokas M, Martin KJ, et al. Detection of cognitive decline after coronary surgery: a comparison of computerized and conventional tests. Br J Anaesth 2004;92(6):814–20. [20] Brott T, Adams HP, Olinger CP, Marler JR, Barson WG, Biller J, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989;20:864–70. [21] Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. A prospective study of acute cerebrovascular disease in the community: the Oxfordshire Community Stroke Project—1981–86. 2. Incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intracerebral and subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry 1990 Jan;53(1):16–22.
106
T.B. Cumming et al. / Journal of the Neurological Sciences 322 (2012) 102–106
[22] Nasreddine ZS, Phillips NA, Bedirian V, Charbonneau S, Whitehead V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 2005;53(4):695–9. [23] Hachinski V, Iadecola C, Petersen RC, Breteler MM, Nyenhuis DL, Black SE, et al. National Institute of Neurological Disorders and Stroke-Canadian Stroke Network vascular cognitive impairment harmonization standards. Stroke 2006;37(9):2220–41.
[24] Middleton LE, Yaffe K. Promising strategies for the prevention of dementia. Arch Neurol 2009;66:1210–5. [25] Fredrickson J, Maruff P, Woodward M, Moore L, Fredrickson A, Sach J, et al. Evaluation of the usability of a brief computerized cognitive screening test in older people for epidemiological studies. Neuroepidemiology 2010;34:65–75.
Contents lists available at SciVerse ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Cutting a long story short: Reaction times in acute stroke are associated with longer term cognitive outcomes Toby B. Cumming a, b,⁎, Amy Brodtmann a, b, David Darby b, c, Julie Bernhardt a a b c
Stroke Division, Florey Neuroscience Institutes, Melbourne, Australia Cognitive Neuroscience Division, Florey Neuroscience Institutes, Melbourne, Australia Department of Medicine, University of Melbourne, Australia
a r t i c l e
i n f o
Article history: Received 31 January 2012 Received in revised form 4 June 2012 Accepted 2 July 2012 Available online 20 July 2012 Keywords: Cerebrovascular disease Cognitive impairment Attention Psychomotor Neuropsychology
a b s t r a c t Background: The viability and usefulness of cognitive assessment in acute stroke have been questioned, with practical challenges arising from the focal nature of neurological deficits as well as heterogeneity in arousal state. We aimed to test whether acute measures of attention correlate with attentional function at 3 months post-stroke. Methods: Patients with confirmed stroke completed 2 computerised cognitive tasks (CogState) within 2 weeks of stroke. The tasks were a simple reaction time task (Detection) and a choice reaction time task (Identification) that required a button press to visual stimuli (playing cards). Each task took approximately 4 min. The Montreal Cognitive Assessment (MoCA) and an extended neuropsychological battery were administered at 3 months post-stroke. Results: Thirty-three patients (mean age 75.5 years, SD 11.9) participated in this preliminary study. Correlations indicated that both Detection speed (r=−0.73, p b 0.001) and Identification speed (r=−0.61, p =0.007) at baseline were associated with attentional function at 3 months, as measured by established neuropsychological tests (Trails-A, Digit span, Digit symbol). In addition, Detection speed at baseline was correlated with total 3-month MoCA score (r=−0.54, p=0.012). Conclusion: Simple and brief computerised assessment of attentional function in acute stroke is possible and is related to longer term attentional and cognitive performance. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Approximately two thirds of stroke survivors experience cognitive impairment [1], with many going on to develop dementia [2]. Stroke appears to have particularly strong effects in certain cognitive domains, with impairments being most pronounced in the areas of attention/executive function and speed of processing [3]. As cognitive domains are not separate but inter-dependent, deficits in one area can influence performance in other areas. For example, the ability to direct, shift and sustain attention underpins many mental skills, including aspects of memory [4]. Thus it may be possible to cut the long story of multi-domain neuropsychological testing short. While the broad classification of ‘cognitive impairment’ after stroke has been associated with increased death and disability [5], greater rates of institutionalisation [6] and poorer rehabilitation outcomes [7], processing speed alone appears related to functional outcome after stroke [8]. Early identification of post-stroke cognitive deficits has the potential to guide therapy (e.g., ‘attention process training’ for attentional deficits ⁎ Corresponding author at: Melbourne Brain Centre, 245 Burgundy St, Heidelberg, 3084, Australia. Tel.: +61 390357152; fax: +61 394962251. E-mail address: [email protected] (T.B. Cumming). 0022-510X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2012.07.004
[9]) and inform rehabilitation strategies. There are practical barriers, however, to assessing cognition early after stroke. Many acute stroke patients have reduced level of consciousness, dysphasia, visual deficits and weakness in the hand and arm, all of which can limit their ability to complete cognitive assessments. In one study, researchers attempted to demonstrate the feasibility of neuropsychological testing between 4 and 20 days after first ischaemic stroke. Yet of the 73 included patients, only 57 were examined, and only 44 could complete 82% of the tests [10]. Shorter screening tools such as the Mini-Mental State Examination and the Montreal Cognitive Assessment have been employed in acute stroke [11,12], although their reliance on written and verbal responses means many moderate and severe stroke patients cannot complete them. There is also some doubt about whether cognitive status in the often medically unstable acute stage accurately reflects cognitive or functional status in the longer term. Early cognitive deficits have been related to longer term cognitive impairment, dependence in activities of daily living, depression and lower quality of life [13,14], but more evidence is required. It is possible that simple computerised cognitive tasks can overcome some of the difficulties in the assessment of acute stroke patients. Computerised reaction time tasks have been used in stroke samples to test attention, processing speed and working memory [15,16], but
T.B. Cumming et al. / Journal of the Neurological Sciences 322 (2012) 102–106
103
2. Methods
flipped over. The card was either a red or a black joker. The task ended after 30 correct trials were completed. Visual and auditory feedback differentiated correct and incorrect responses, and anticipatory responses (button press before card flipped over) triggered a distinctive error sound. Accuracy and reaction times for correct responses were recorded. The MoCA [22] is scored out of 30 and includes items on visuospatial ability, executive function, naming, memory, attention, language, abstract thinking and orientation. It has been recommended as a useful screen for vascular cognitive impairment [23], and appears to be slightly more sensitive than the Mini-Mental State Examination after stroke [12]. The attentional tasks administered as part of the 3-month neuropsychological battery were Trail-Making part A (draw a line between numbers 1 to 25 in ascending order), Digit Span (repeat number strings) and Digit Symbol (match symbols to numbers according to a specified code). Z-scores were derived for each patient on each of these 3 tasks using well-established age-based population norms. These 3 z-scores were then averaged to form a single summed z-score for attentional function.
2.1. Participants
2.4. Data analysis
Patients who were admitted to the acute stroke unit at the Austin Hospital with completed stroke (ischaemic or intracerebral haemorrhage) were eligible for this study. Patients were excluded if they: (a) were younger than 18 years old, (b) suffered the stroke more than 2 weeks previously, (c) required an interpreter, or (d) had major visual, hearing or receptive language impairments which prevented completion of the cognitive tasks. All patients provided informed consent prior to inclusion in the study. Relevant demographic and clinical characteristics, including age, stroke severity using the NIHSS [20], Oxfordshire classification [21] and lesion side, were extracted from the patient's medical record. Ethical approval for this study was obtained from the Austin Health Human Research Ethics Committee.
For the computerised tasks, a base 10 logarithmic transformation was applied to normalise the distributions of mean reaction time. If a patient made an error on 50% or more of the trials, their reaction time data for that task were excluded from analysis. Pearson correlations were computed between acute reaction times and a composite 3-month attentional score. Multivariate linear regressions were then conducted that included age and baseline stroke severity (NIHSS) as confounding variables. Additional correlations were performed to test the relationship between baseline stroke severity (NIHSS) and the acute reaction time data. As there were no previous data on which to base an estimate of effect size, no formal sample size calculation was performed. We aimed to include 30 patients to justify including 3 independent variables in regression, so attempted to recruit 33 to allow for a 10% dropout rate.
typically not in the acute stage. The CogState computerised battery (CogState Ltd, Melbourne, Australia) is composed of tasks that were derived from the cognitive paradigms of psychomotor function, attention and working memory. These reaction time tasks are short, straightforward and not reliant on expressive language; they require button press responses to the presentation of playing cards on a computer screen. Research in other populations has shown the CogState tasks to be valid and reliable [17], free of practice effects in serial testing [18] and sensitive to change [19]. In this preliminary study, we hypothesised that response times on simple and choice reaction time tasks in the acute stage (within 2 weeks) of stroke would be correlated with standard measures of attention at 3 months post-stroke. A second hypothesis was that acute reaction times would be correlated with total score on the Montreal Cognitive Assessment (MoCA) at 3 months.
2.2. Procedure 3. Results The baseline testing session took place on the acute stroke ward immediately following written informed consent. The cognitive tasks were presented on a laptop computer that had 2 large (approximately 6 cm in diameter) external response keys (the ‘yes’ and ‘no’ buttons) attached. Each patient was either sitting up in bed or sitting in a chair next to their bed for the testing (wherever they were originally placed). Curtains were drawn to minimise interruptions and to encourage the patient to focus on the task. At 3 months post-stroke, 2 follow-up visits were made to the patient's residence. During the first visit, a cognitive screening tool (MoCA) was completed. At the second visit, a comprehensive neuropsychological battery was administered that included attentional tasks. The same researcher (TBC) administered the neuropsychological battery and the acute testing session, but he remained blind to the results of the acute cognitive tasks until the end of the study. 2.3. Outcome measures The computerised tasks administered in the acute stage were the Detection task and the Identification task from the CogState battery. Both tasks were preceded by instructions read aloud by the experimenter and a series of practice items for task familiarisation. A playing card was presented face down in the centre of the screen and, after an interval that varied randomly between 2.5 and 3.5 s, instantly turned face up. The Detection task is a simple reaction time test of psychomotor function and speed of processing; the person was instructed to press the ‘yes’ button as soon as possible after the card on screen flipped over. The card was always the same joker card. The task ended after 35 correct trials were completed. The Identification task is a choice reaction time test of visual attention; the person was instructed to press the ‘yes’ button if the card is red and the ‘no’ button if the card is not red as soon as possible after it
Thirty-three patients participated; demographics and stroke characteristics are shown in Table 1. The predominantly male sample had a mean age of 75.5 and a mean acute NIHSS score of 5.1. 3.1. Feasibility Baseline testing was completed at a mean of 5.4 days post-stroke (SD =2.9), with a range between 1 and 12 days. Each of the computerised tasks can be completed within 3 min, including instructions and practice items. A number of stroke patients took longer to understand the instructions, made a number of errors and had slowed responses, so these tasks often took approximately 4 min each. For the Detection task, 32 patients had complete data. One patient finished the task, but their data could not be retrieved due to equipment failure. Six patients did not get more than 50% correct and were therefore excluded from calculations, leaving data from 26. For the Identification task, 31 patients had complete data. One patient finished the task, but their data could not be retrieved due to equipment failure. Another patient had to stop testing after the Detection task due to dizziness and weakness, and thus did not complete the Identification task. Seven patients failed to perform above chance (50%) and were therefore excluded from calculations, leaving data from 24. Log10 reaction times are presented in Table 2. Hypothesis 1. Detection speed at baseline was significantly correlated with attentional function z-score at 3 months (n =21; r=−0.73, pb 0.001). This relationship is illustrated in Fig. 1. In multivariable regression, Detection speed at baseline was significantly associated with attentional function at 3 months (t=−5.24, pb 0.001), with age (p=0.15) and stroke severity (p=0.07) included as independent variables. This
104
T.B. Cumming et al. / Journal of the Neurological Sciences 322 (2012) 102–106
Table 1 Patient characteristics (N = 33). N (%)a Age — mean (SD), range Male NIHSS score — mean (SD), range Mild (1–7) Moderate (8–15) Oxfordshire stroke classification TACI PACI POCI LACI ICH Left-sided lesion Right-sided lesion
75.5 (11.9), 45–95 25 (76) 5.1 (3.2), 1–15 26 (79) 7 (21) 1 (3) 18 (55) 9 (27) 3 (9) 2 (6) 16 (52) 15 (48)
NIHSS — National Institutes of Health Stroke Scale; TACI — total anterior circulation infarct; PACI — partial anterior circulation infarct; POCI – posterior circulation infarct; LACI — lacunar infarct; ICH — intracerebral haemorrhage. a N (%) unless otherwise specified. Lesion side missing for 2 patients. Fig. 1. Relationship between reaction time on the Detection (DET) task at baseline and summed 3-month attention z-score. 2
model was highly significant (R =0.66; F=10.8, pb 0.001). Identification speed at baseline was also correlated with attentional function at 3 months (n=18; r=−0.61, p=0.007). In regression, Identification speed at baseline was significantly related to 3-month attentional function (t = − 3.50, p = 0.004), with age (p = 0.12) and stroke severity (p = 0.59) included. This regression model was also significant (R 2 = 0.51; F = 4.8, p = 0.017). Hypothesis 2. Detection speed at baseline was significantly correlated with MoCA score at 3 months (n = 21; r = − 0.54, p = 0.012). Fig. 2 shows this relationship. In multivariable regression, baseline Detection speed was significantly associated with MoCA score at 3 months (t = − 2.39, p = 0.029), with age (p = 0.11) and stroke severity (p = 0.61) included. This regression model was significant (R 2 = 0.40; F = 3.7, p = 0.031). Identification speed at baseline was moderately correlated with MoCA score at 3 months, but this did not reach significance (n = 18; r = − 0.39, p = 0.11). In multivariable regression, Identification speed at baseline was not significantly related to MoCA at 3 months (t=−1.27, p=0.23), with age (p=0.14) and stroke severity (p=0.75) included. This regression model was not significant (R2 =0.30; F=2.0, p=0.16). There was no significant correlation between acute stroke severity (NIHSS) and Detection speed (n= 26; r = 0.04, p = 0.84; see Fig. 3) or Identification speed (n= 24; r = 0.18, p = 0.41) at baseline.
the first 3 weeks after stroke and longer term cognitive impairment. This prior study, however, employed standard neuropsychological testing in both the acute setting and long-term follow-up. Our findings indicate that simple and choice reaction time tasks administered in the first week following stroke can also be prognostic for longer term cognitive performance. One possible explanation is that acute reaction times were merely a proxy for stroke severity. This is unlikely, however, given the lack of correlation between acute NIHSS score and speed on the Detection and Identification tasks. We were surprised to find no relationship here, as stroke severity is normally a strong predictor of not only cognitive but other functional outcomes. It appears that the computerised cognitive tasks provide important information beyond a clinical impression of neurological impairment. Score on the NIHSS scale, which incorporates a wide range of neurological deficits including level of consciousness, dysphasia, neglect, visual field loss, motor weakness, ataxia, dysarthria and sensory loss, does not necessarily account for differences in attention and speed of processing. There is growing awareness of the role that stroke and vascular risk factors play in the development of cognitive impairment and dementia [2,24]. Ideally, there would be cognitive assessments that could be employed soon after stroke (to obtain a post-stroke baseline) and that could be repeated multiple times subsequently (to track recovery or decline). The Detection and Identification tasks used here have these
4. Discussion We found that the reaction times of acute stroke patients on simple computerised tasks were related to attentional function assessed at 3 months post-stroke. These correlations could not be explained by age or stroke severity, as the associations remained when these variables were accounted for in multivariable regression. We also documented a significant relationship between speed on the Detection task (simple reaction time) at baseline and total score on the MoCA (a global cognitive screening tool) at 3 months post-stroke. There is some precedent for these results, with Nys et al. [13] reporting a strong relationship between domain-specific cognitive abilities tested in
Table 2 Log10 reaction times (and millisecond equivalents) for the Detection (DET) and Identification (IDN) tasks at baseline.
DET IDN
N
Mean log10 RT
SD
Min
Max
Mean ms
26 24
2.64 2.83
0.17 0.12
2.37 2.54
2.98 3.05
437 679
Fig. 2. Relationship between reaction time on the Detection (DET) task at baseline and total MoCA score at 3 months.
T.B. Cumming et al. / Journal of the Neurological Sciences 322 (2012) 102–106
105
Acknowledgements We thank all participants for their time and effort. We also thank Debbie Hansen and Karen Moss for helping collect the data. This work was supported by research grants from the National Stroke Foundation and from the Equity Trustees Preston & Loui Geduld Trust Fund. Dr Cumming is funded by a National Heart Foundation Postdoctoral Research Fellowship. All research undertaken at the Florey Neuroscience Institutes is supported by infrastructure funding from the Department of Innovation (DOBI).
References
Fig. 3. Relationship between reaction time on the Detection (DET) task at baseline and stroke severity (NIHSS score) at baseline.
desirable characteristics. As straightforward reaction time tasks, they do not cover the range of cognitive domains that a neuropsychological battery covers. They may be very useful, though, as measures of attentional function and possibly therefore as surrogate markers for general cognition. The relationship between lesion location and profile of cognitive deficits is not always predictable, and impairment may arise separate to any focal injury directly attributable to the stroke. It may be that infarction anywhere in the brain results in disruption to the widely distributed attention network – or to its critical and abundant neurotransmitters – which is then reflected in slower processing, more diffuse activation and more disorganised thinking. Simple and choice reaction times might be markers for the extent of this attention network disruption. As a group, the acute stroke patients tested here exhibited slower reaction times than a group of healthy older people tested on the same Detection and Identification tasks. Fredrickson et al. [25] reported data for a group of 301 community-dwelling people with mean age of 62. The mean age in our stroke sample was 76. Although some slowing would be expected purely due to age, the patients in our study were markedly slower. With log10 reaction times translated to milliseconds, the stroke patients took longer than the community sample in Detection (mean 437 ms versus 331 ms) and Identification (mean 679 ms versus 525 ms). Several limitations to the current study should be noted. The sample was small, and as such our results require replication in larger groups. The sample was made up of mild and moderate stroke patients only, and our findings therefore lack generalisability to more severe patients. While the CogState tasks appear more feasible than many other cognitive assessment options, there are still acute stroke patients whose level of consciousness or receptive dysphasia makes completing the tasks impossible. The patients were not sampled consecutively, although this should not pose a major problem as the analyses presented here are mostly within-subject. Finally, we have not examined the connection between lesion location and task performance, and therefore can make no conclusions about the influence of stroke topography. Our results indicate that simple and brief computerised assessment of attentional function in acute stroke is possible and is related to longer term attentional and cognitive performance. The exciting prospect that reaction time tasks can be used not only to track attentional performance over time following stroke but also as surrogate markers for more general cognitive function requires further evaluation in larger studies.
Conflict of interest Dr. Darby is a shareholder and consultant for CogState Ltd.
[1] Jin YP, Di Legge S, Ostbye T, Feightner JW, Hachinski V. The reciprocal risks of stroke and cognitive impairment in an elderly population. Alzheimers Dement 2006;2:171–8. [2] Pendlebury ST, Rothwell PM. Prevalence, incidence, and factors associated with pre-stroke and post-stroke dementia: a systematic review and meta-analysis. Lancet Neurol 2009 Nov;8(11):1006–18. [3] Feigin VL, Barker-Collo S, Parag V, Senior H, Lawes CM, Ratnasabapathy Y, et al. Auckland Stroke Outcomes Study. Part 1: gender, stroke types, ethnicity, and functional outcomes 5 years poststroke. Neurology 2010;75(18):1597–607. [4] McNab F, Klingberg T. Prefrontal cortex and basal ganglia control access to working memory. Nat Neurosci 2008;11:103–7. [5] Patel MD, Coshall C, Rudd AG, Wolfe CD. Cognitive impairment after stroke: clinical determinants and its associations with long-term stroke outcomes. J Am Geriatr Soc 2002;50(4):700–6. [6] Pasquini M, Leys D, Rousseaux M, Pasquier F, Henon H. Influence of cognitive impairment on the institutionalisation rate 3 years after a stroke. J Neurol Neurosurg Psychiatry 2007;78(1):56–9. [7] Heruti RJ, Lusky A, Dankner R, Ring H, Dolgopiat M, Barell V, et al. Rehabilitation outcome of elderly patients after a first stroke: effect of cognitive status at admission on the functional outcome. Arch Phys Med Rehabil 2002;83(6):742–9. [8] Barker-Collo S, Feigin VL, Parag V, Lawes CM, Senior H. Auckland Stroke Outcomes Study. Part 2: cognition and functional outcomes 5 years poststroke. Neurology 2010;75(18):1608–16. [9] Barker-Collo SL, Feigin VL, Lawes CM, Parag V, Senior H, Rodgers A. Reducing attention deficits after stroke using attention process training: a randomized controlled trial. Stroke 2009;40(10):3293–8. [10] van Zandvoort MJ, Kessels RP, Nys GM, de Haan EH, Kappelle LJ. Early neuropsychological evaluation in patients with ischaemic stroke provides valid information. Clin Neurol Neurosurg 2005;107(5):385–92. [11] Dong Y, Sharma VK, Chan BP, Venketasubramanian N, Teoh HL, Seet RC, et al. The Montreal Cognitive Assessment (MoCA) is superior to the Mini-Mental State Examination (MMSE) for the detection of vascular cognitive impairment after acute stroke. J Neurol Sci 2010;299(1–2):15–8. [12] Godefroy O, Fickl A, Roussel M, Auribault C, Bugnicourt JM, Lamy C, et al. Is the Montreal Cognitive Assessment Superior to the Mini-Mental State Examination to detect poststroke cognitive impairment?: a study with neuropsychological evaluation. Stroke 2011;42(6):1712–6 (2011 June 1). [13] Nys GM, van Zandvoort MJ, de Kort PL, van der Worp HB, Jansen BP, Algra A, et al. The prognostic value of domain-specific cognitive abilities in acute first-ever stroke. Neurology 2005;64(5):821–7. [14] Nys GM, van Zandvoort MJ, van der Worp HB, de Haan EH, de Kort PL, Jansen BP, et al. Early cognitive impairment predicts long-term depressive symptoms and quality of life after stroke. J Neurol Sci 2006;247(2):149–56. [15] Ballard C, Stephens S, Kenny R, Kalaria R, Tovee M, O'Brien J. Profile of neuropsychological deficits in older stroke survivors without dementia. Dement Geriatr Cogn Disord 2003;16(1):52–6. [16] Quaney BM, Boyd LA, McDowd JM, Zahner LH, He J, Mayo MS, et al. Aerobic exercise improves cognition and motor function poststroke. Neurorehabil Neural Repair 2009 Nov;23(9):879–85. [17] Collie A, Maruff P, Makdissi M, McCrory P, McStephen M, Darby D. CogSport: reliability and correlation with conventional cognitive tests used in postconcussion medical evaluations. Clin J Sport Med 2003;13(1):28–32. [18] Falleti MG, Maruff P, Collie A, Darby DG. Practice effects associated with the repeated assessment of cognitive function using the CogState battery at 10-minute, one week and one month test-retest intervals. J Clin Exp Neuropsychol Off J Int Neuropsychol Soc 2006;28(7):1095–112. [19] Silbert BS, Maruff P, Evered LA, Scott DA, Kalpokas M, Martin KJ, et al. Detection of cognitive decline after coronary surgery: a comparison of computerized and conventional tests. Br J Anaesth 2004;92(6):814–20. [20] Brott T, Adams HP, Olinger CP, Marler JR, Barson WG, Biller J, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989;20:864–70. [21] Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. A prospective study of acute cerebrovascular disease in the community: the Oxfordshire Community Stroke Project—1981–86. 2. Incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intracerebral and subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry 1990 Jan;53(1):16–22.
106
T.B. Cumming et al. / Journal of the Neurological Sciences 322 (2012) 102–106
[22] Nasreddine ZS, Phillips NA, Bedirian V, Charbonneau S, Whitehead V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 2005;53(4):695–9. [23] Hachinski V, Iadecola C, Petersen RC, Breteler MM, Nyenhuis DL, Black SE, et al. National Institute of Neurological Disorders and Stroke-Canadian Stroke Network vascular cognitive impairment harmonization standards. Stroke 2006;37(9):2220–41.
[24] Middleton LE, Yaffe K. Promising strategies for the prevention of dementia. Arch Neurol 2009;66:1210–5. [25] Fredrickson J, Maruff P, Woodward M, Moore L, Fredrickson A, Sach J, et al. Evaluation of the usability of a brief computerized cognitive screening test in older people for epidemiological studies. Neuroepidemiology 2010;34:65–75.