Clinical Neurophysiology 126 (2015) 689–697
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
P300 amplitude and response speed relate to preserved cognitive function in relapsing–remitting multiple sclerosis Mathias Sundgren a, Vadim V. Nikulin b,c, Liselotte Maurex a, Åke Wahlin d, Fredrik Piehl a, Tom Brismar a,⇑ a
Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden Neurophysics Group, Department of Neurology, Campus Benjamin Franklin, Charité – University Medicine Berlin, Germany c Bernstein Center for Computational Neuroscience, Berlin, Germany d Institute of Gerontology, School of Health Sciences, Jönköping University, Jönköping, Sweden b
a r t i c l e
i n f o
Article history: Accepted 14 July 2014 Available online 20 August 2014 Keywords: Multiple sclerosis Event-related potential P300 Response time Cognitive function
h i g h l i g h t s P300 amplitudes and response times in choice reaction tasks (visual and auditory stimulation) have
strong association with cognitive function in relapsing–remitting MS which differs from the absent or weak association in healthy controls. These associations were not an epiphenomenon of the cognitive decline in patients, since parietal P300 amplitudes and response times were normal. Patients had an increase in the visual P300 amplitude in the frontal region suggesting activation of compensatory networks.
a b s t r a c t Objective: To explore if cognitive impairment in relapsing–remitting multiple sclerosis (RRMS) is associated with abnormal neural function and if there is evidence of neural compensatory mechanisms. Methods: Seventy-two RRMS patients and 89 healthy control subjects were included in a cross-sectional study. Event-related brain potential (P300) and response time (RT) were recorded with visual and auditory choice reaction tasks. Cognitive function was evaluated with an 18 item test battery. Results: Patients had a decrease in cognitive function (p < 0.001 for global score) and increased visual P300 amplitude frontally. P300 amplitude was normal in other brain areas and RT was normal. P300 latency was normal except for an increase in auditory latency occipitally. Cognitive performance correlated positively with parietal P300 amplitude in patients but not in controls. Cognition had stronger correlation (negative) with RT in patients than in controls. Conclusions: Patients with low P300 amplitude and long RT were more often cognitively impaired. This indicates that general factors such as signal amplitude and speed are limiting for cognitive function in RRMS patients. The increase in frontal P300 amplitude may be a compensatory effect. Significance: Our findings suggest that high amplitude and fast speed may be protective against cognitive impairment. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Cognitive impairment is frequent in multiple sclerosis (MS) and appears in all subtypes of the disease (Potagas et al., 2008). It cor⇑ Corresponding author. Address: Clinical Neurophysiology, Karolinska University Hospital (Solna) R2:01, SE-17176 Stockholm, Sweden. Tel.: +46 851772030; fax: +46 851774415. E-mail address: [email protected] (T. Brismar).
relates with brain MRI metrics, especially measures of grey matter and brain atrophy, but the explained variance remains moderate (Benedict and Zivadinov, 2011). This imperfect correlation between measurable neuropathology and degree of cognitive impairment is not unique for MS and has been well documented in Alzheimer’s disease (AD) (Stern, 2006). High premorbid intelligence has been proposed to be a moderating factor in ageing and AD (see Stern, 2009) and a similar protective effect has been reported in MS (see Sumowski and Leavitt, 2013).
http://dx.doi.org/10.1016/j.clinph.2014.07.024 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
690
M. Sundgren et al. / Clinical Neurophysiology 126 (2015) 689–697
Several functional MRI (fMRI) studies have indicated that cognitive task performance is associated with increased or altered cortical activation patterns in patients with MS. During a processing speed and working memory task patients with relapsing–remitting MS (RRMS) activated larger brain areas located in the bilateral prefrontal and inferior parietal cortex than controls (Mainero et al., 2004). Staffen et al. (2012) have pointed out that brain activation during stimulation with difficult cognitive tests may be confounded by compensatory mechanisms related to premorbid intelligence. Therefore they used a simple test such as the oddball paradigm for stimulation in order to avoid this effect. With this paradigm they found that patients with mild cognitive impairment (MCI) displayed a widespread decrease in fMRI activation as compared to controls. This is consistent with event-related potential (ERP) studies demonstrating that the electrical P300 response elicited by the oddball paradigm is decreased in MCI (Parra et al., 2012; van Deursen et al., 2009). The P300 component is associated with the cognitive processes of volitional target detection and is generated over widespread cortical regions in healthy subjects (Juckel et al., 2012). The P300 amplitude increases in proportion to the amount of attentional resources devoted to a given task. Any brain disorder affecting cognitive processes may reduce the P300 amplitude and increase the latency (see Polich and Herbst, 2000). Visual stimulation tasks requiring feature analysis were found to elicit an increase in the amplitude of the earlier components (in the latency range 150–300 ms) in the fronto-central region (Luck and Hillyard, 1994; Luck et al., 2000). In early MS Pelosi et al. (1997) found that the early positive peak (P145) was delayed in a visual stimulation working memory task. Most previous ERP-studies have included mixed samples of MSpatients, including both cognitively less affected patients with RRMS as well as more affected patients with progressive subtypes of the disease, making inferences or generalizations regarding subgroups of MS difficult. RRMS is the largest subgroup and constitutes an estimated 58% of the entire MS population (Confavreux and Vukusic, 2006). Studies including ERP recordings in different subgroups of MS have generally reported reduced amplitude and increased latency of the P300 component (see Leocani et al., 2010; Magnano et al., 2006). These ERP studies include auditory, visual or both modalities. Larger effects on P300 amplitude and latency are seen in the progressive subtypes as compared to RRMS (Ellger et al., 2002). Furthermore, the visual and auditory P300 amplitude has a negative correlation with processing speed measured over a 12 month period in MS patients (subgroups not specified) (Kiiski et al., 2011). Speed of volitional target detection can be assessed with reaction or response time (RT) tasks. Fast RT is related to high general cognitive ability in healthy individuals (Deary et al., 2001). There is a similar correlation between RT and processing speed in RRMS patients (Tombaugh et al., 2010). They found that a choice RT task is a more sensitive measure of impaired information processing in RRMS as compared to a simple RT task. The purpose of the present study was to explore if cognitive impairment in RRMS is associated with abnormal neuronal function, and if there is evidence of neural compensatory mechanisms. The simplest hypothesis to test was that cognitive deficits are associated with reduced P300 amplitude, increased latency and increased RT. However, as pointed out above some factors may modify this effect. The P300 amplitude may increase due to an increase in attentional resources or effort. Another hypothesis was that the association between cognitive function and ERP measures differs in patients and control subjects. This may be due to disease related alterations in the function in the underlying neural networks. The study was restricted to RRMS in order to eliminate the variability related to differences between subgroups. The target P300
response and RT were determined with choice reaction tasks and performed with both auditory and visual stimulation in order to capture modality specific effects. The analysis of the visual ERP also included positive components earlier than P300 which may be associated to cognitive processes. The specific aims were to examine (1) if RRMS patients display abnormalities in the P300 amplitude and RT, and (2) if cognitive domain functions have a different relation to P300 amplitude and RT in patients compared to controls. 2. Methods 2.1. Subjects Seventy-two patients with RRMS and 89 healthy control subjects were examined. The subjects, clinical instruments and neuropsychological tests were the same as described in a previous study (Sundgren et al., 2013). The patient and control groups did not differ significantly in age, sex ratio, handedness or education (Table 1). The mean accumulated physical disability was mild. Most patients had disease modifying treatment (DMT), the most frequent being natalizumab (36%) followed by interferon-beta (33%), glatiramer acetate (8%), or unknown (clinical trial, 8%). Fourteen percent had no DMT. All subjects were informed about the nature and purpose of the study before consenting to participate. The protocol was approved by the regional ethics committee (Regionala etikprövningsnämnden i Stockholm). The study was conducted in accordance with Good Clinical Practice guidelines and the principles of the Declaration of Helsinki. 2.2. Clinical instruments All patients underwent a clinical neurological examination. Physical disability was assessed by the Kurtzke Expanded Disability Status Scale (EDSS) (Kurtzke, 1983) and The Multiple Sclerosis Severity Score (MSSS) (Roxburgh et al., 2005) was used for assessment of disease severity. Symptoms of depression were assessed by the Beck Depression Inventory (BDI) (Beck et al., 1988). Fatigue was assessed with the Fatigue Severity Scale (FSS) (Krupp et al., 1989). Visual acuity was tested with Snellen charts in 64 patients and in all control subjects. Seven patients and 7 control subjects had visual acuity less than 1.0. Poor vision that could interfere with the testing procedures was an exclusion criteria (Sundgren et al., 2013). Hearing thresholds were not tested in the patients. No participant had signs of hearing impairment during the examinations. 2.3. Neuropsychological tests and cognitive domain functions Cognitive function was evaluated with nine neuropsychological tests resulting in a total of 18 cognitive test scores (Table 2). The
Table 1 Demographic and clinical data of study population.
Age (years) Female sex (%) Left handedness (%) Education (years) Disease duration (years) EDSS (range 0–10) MSSS (range 0.01–9.99) BDI (range 0–63) FSS (range 1–7)
Patients (N = 72)
Control subjects (N = 89)
Mean
S.D.
Mean
S.D.
37.9 71 12.5 13.8 9.3 2.7 4.1 8.8 3.9
10.0
38.2 57 9.0 14.1
11.5
2.5
n.s. n.s. (0.08) n.s. n.s.
4.0 2.6
4.2 1.0
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
P300 amplitude and response speed relate to preserved cognitive function in relapsing–remitting multiple sclerosis Mathias Sundgren a, Vadim V. Nikulin b,c, Liselotte Maurex a, Åke Wahlin d, Fredrik Piehl a, Tom Brismar a,⇑ a
Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden Neurophysics Group, Department of Neurology, Campus Benjamin Franklin, Charité – University Medicine Berlin, Germany c Bernstein Center for Computational Neuroscience, Berlin, Germany d Institute of Gerontology, School of Health Sciences, Jönköping University, Jönköping, Sweden b
a r t i c l e
i n f o
Article history: Accepted 14 July 2014 Available online 20 August 2014 Keywords: Multiple sclerosis Event-related potential P300 Response time Cognitive function
h i g h l i g h t s P300 amplitudes and response times in choice reaction tasks (visual and auditory stimulation) have
strong association with cognitive function in relapsing–remitting MS which differs from the absent or weak association in healthy controls. These associations were not an epiphenomenon of the cognitive decline in patients, since parietal P300 amplitudes and response times were normal. Patients had an increase in the visual P300 amplitude in the frontal region suggesting activation of compensatory networks.
a b s t r a c t Objective: To explore if cognitive impairment in relapsing–remitting multiple sclerosis (RRMS) is associated with abnormal neural function and if there is evidence of neural compensatory mechanisms. Methods: Seventy-two RRMS patients and 89 healthy control subjects were included in a cross-sectional study. Event-related brain potential (P300) and response time (RT) were recorded with visual and auditory choice reaction tasks. Cognitive function was evaluated with an 18 item test battery. Results: Patients had a decrease in cognitive function (p < 0.001 for global score) and increased visual P300 amplitude frontally. P300 amplitude was normal in other brain areas and RT was normal. P300 latency was normal except for an increase in auditory latency occipitally. Cognitive performance correlated positively with parietal P300 amplitude in patients but not in controls. Cognition had stronger correlation (negative) with RT in patients than in controls. Conclusions: Patients with low P300 amplitude and long RT were more often cognitively impaired. This indicates that general factors such as signal amplitude and speed are limiting for cognitive function in RRMS patients. The increase in frontal P300 amplitude may be a compensatory effect. Significance: Our findings suggest that high amplitude and fast speed may be protective against cognitive impairment. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Cognitive impairment is frequent in multiple sclerosis (MS) and appears in all subtypes of the disease (Potagas et al., 2008). It cor⇑ Corresponding author. Address: Clinical Neurophysiology, Karolinska University Hospital (Solna) R2:01, SE-17176 Stockholm, Sweden. Tel.: +46 851772030; fax: +46 851774415. E-mail address: [email protected] (T. Brismar).
relates with brain MRI metrics, especially measures of grey matter and brain atrophy, but the explained variance remains moderate (Benedict and Zivadinov, 2011). This imperfect correlation between measurable neuropathology and degree of cognitive impairment is not unique for MS and has been well documented in Alzheimer’s disease (AD) (Stern, 2006). High premorbid intelligence has been proposed to be a moderating factor in ageing and AD (see Stern, 2009) and a similar protective effect has been reported in MS (see Sumowski and Leavitt, 2013).
http://dx.doi.org/10.1016/j.clinph.2014.07.024 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
690
M. Sundgren et al. / Clinical Neurophysiology 126 (2015) 689–697
Several functional MRI (fMRI) studies have indicated that cognitive task performance is associated with increased or altered cortical activation patterns in patients with MS. During a processing speed and working memory task patients with relapsing–remitting MS (RRMS) activated larger brain areas located in the bilateral prefrontal and inferior parietal cortex than controls (Mainero et al., 2004). Staffen et al. (2012) have pointed out that brain activation during stimulation with difficult cognitive tests may be confounded by compensatory mechanisms related to premorbid intelligence. Therefore they used a simple test such as the oddball paradigm for stimulation in order to avoid this effect. With this paradigm they found that patients with mild cognitive impairment (MCI) displayed a widespread decrease in fMRI activation as compared to controls. This is consistent with event-related potential (ERP) studies demonstrating that the electrical P300 response elicited by the oddball paradigm is decreased in MCI (Parra et al., 2012; van Deursen et al., 2009). The P300 component is associated with the cognitive processes of volitional target detection and is generated over widespread cortical regions in healthy subjects (Juckel et al., 2012). The P300 amplitude increases in proportion to the amount of attentional resources devoted to a given task. Any brain disorder affecting cognitive processes may reduce the P300 amplitude and increase the latency (see Polich and Herbst, 2000). Visual stimulation tasks requiring feature analysis were found to elicit an increase in the amplitude of the earlier components (in the latency range 150–300 ms) in the fronto-central region (Luck and Hillyard, 1994; Luck et al., 2000). In early MS Pelosi et al. (1997) found that the early positive peak (P145) was delayed in a visual stimulation working memory task. Most previous ERP-studies have included mixed samples of MSpatients, including both cognitively less affected patients with RRMS as well as more affected patients with progressive subtypes of the disease, making inferences or generalizations regarding subgroups of MS difficult. RRMS is the largest subgroup and constitutes an estimated 58% of the entire MS population (Confavreux and Vukusic, 2006). Studies including ERP recordings in different subgroups of MS have generally reported reduced amplitude and increased latency of the P300 component (see Leocani et al., 2010; Magnano et al., 2006). These ERP studies include auditory, visual or both modalities. Larger effects on P300 amplitude and latency are seen in the progressive subtypes as compared to RRMS (Ellger et al., 2002). Furthermore, the visual and auditory P300 amplitude has a negative correlation with processing speed measured over a 12 month period in MS patients (subgroups not specified) (Kiiski et al., 2011). Speed of volitional target detection can be assessed with reaction or response time (RT) tasks. Fast RT is related to high general cognitive ability in healthy individuals (Deary et al., 2001). There is a similar correlation between RT and processing speed in RRMS patients (Tombaugh et al., 2010). They found that a choice RT task is a more sensitive measure of impaired information processing in RRMS as compared to a simple RT task. The purpose of the present study was to explore if cognitive impairment in RRMS is associated with abnormal neuronal function, and if there is evidence of neural compensatory mechanisms. The simplest hypothesis to test was that cognitive deficits are associated with reduced P300 amplitude, increased latency and increased RT. However, as pointed out above some factors may modify this effect. The P300 amplitude may increase due to an increase in attentional resources or effort. Another hypothesis was that the association between cognitive function and ERP measures differs in patients and control subjects. This may be due to disease related alterations in the function in the underlying neural networks. The study was restricted to RRMS in order to eliminate the variability related to differences between subgroups. The target P300
response and RT were determined with choice reaction tasks and performed with both auditory and visual stimulation in order to capture modality specific effects. The analysis of the visual ERP also included positive components earlier than P300 which may be associated to cognitive processes. The specific aims were to examine (1) if RRMS patients display abnormalities in the P300 amplitude and RT, and (2) if cognitive domain functions have a different relation to P300 amplitude and RT in patients compared to controls. 2. Methods 2.1. Subjects Seventy-two patients with RRMS and 89 healthy control subjects were examined. The subjects, clinical instruments and neuropsychological tests were the same as described in a previous study (Sundgren et al., 2013). The patient and control groups did not differ significantly in age, sex ratio, handedness or education (Table 1). The mean accumulated physical disability was mild. Most patients had disease modifying treatment (DMT), the most frequent being natalizumab (36%) followed by interferon-beta (33%), glatiramer acetate (8%), or unknown (clinical trial, 8%). Fourteen percent had no DMT. All subjects were informed about the nature and purpose of the study before consenting to participate. The protocol was approved by the regional ethics committee (Regionala etikprövningsnämnden i Stockholm). The study was conducted in accordance with Good Clinical Practice guidelines and the principles of the Declaration of Helsinki. 2.2. Clinical instruments All patients underwent a clinical neurological examination. Physical disability was assessed by the Kurtzke Expanded Disability Status Scale (EDSS) (Kurtzke, 1983) and The Multiple Sclerosis Severity Score (MSSS) (Roxburgh et al., 2005) was used for assessment of disease severity. Symptoms of depression were assessed by the Beck Depression Inventory (BDI) (Beck et al., 1988). Fatigue was assessed with the Fatigue Severity Scale (FSS) (Krupp et al., 1989). Visual acuity was tested with Snellen charts in 64 patients and in all control subjects. Seven patients and 7 control subjects had visual acuity less than 1.0. Poor vision that could interfere with the testing procedures was an exclusion criteria (Sundgren et al., 2013). Hearing thresholds were not tested in the patients. No participant had signs of hearing impairment during the examinations. 2.3. Neuropsychological tests and cognitive domain functions Cognitive function was evaluated with nine neuropsychological tests resulting in a total of 18 cognitive test scores (Table 2). The
Table 1 Demographic and clinical data of study population.
Age (years) Female sex (%) Left handedness (%) Education (years) Disease duration (years) EDSS (range 0–10) MSSS (range 0.01–9.99) BDI (range 0–63) FSS (range 1–7)
Patients (N = 72)
Control subjects (N = 89)
Mean
S.D.
Mean
S.D.
37.9 71 12.5 13.8 9.3 2.7 4.1 8.8 3.9
10.0
38.2 57 9.0 14.1
11.5
2.5
n.s. n.s. (0.08) n.s. n.s.
4.0 2.6
4.2 1.0
