European Neuropsychopharmacology (2015) 25, 522–530

www.elsevier.com/locate/euroneuro

Effects of fatigue on cognitive control in neurosarcoidosis Christian Bestea,n,1, Janina Kneiphofa,b,1, Dirk Woitallab a

Cognitive Neurophysiology, Department of Child and Adolescent Psychiatry, Faculty of Medicine of the TU Dresden, Germany b Department of Neurology, St. Josef Hospital, Ruhr-University Bochum, Germany Received 18 August 2014; received in revised form 15 October 2014; accepted 12 January 2015

KEYWORDS

Abstract

Neurosarcoidosis; Response selection; Fatigue; EEG; Cognitive control; Proinflammatory cytokines

Fatigue is a usual reaction to prolonged performance but also a major symptom in various neuroimmunological diseases. In neurosarcoidosis fatigue is a core symptom, but little is known about the relevance of fatigue on cognitive functions in this disease. Previous results in healthy subjects suggest that fatigue strongly affects cognitive control processes. However, fatigue is not a uni-dimensional construct but consists of different facets. It is unknown which of these facets are most important for mechanisms of cognitive control. In the current study we investigate conflict monitoring and response selection processes in neurosarcoidosis patients as a ‘model disease’ of fatigue and healthy controls in relation to the impact of ‘cognitive’ and ‘motor fatigue’ on these processes using event-related potentials (ERPs). We focus on ERPs reflecting attentional selection (P1, N1) and conflict monitoring/response selection processes (N2). ERPs reflecting attentional selection processes were unchanged. The N2 on incompatible trials was reduced in neurosarcoidosis suggesting that response selection and conflict monitoring functions are dysfunctional. Of note, fatigue strongly modulates responses selection processes in conflicting situations (N2) in controls and neurosarcoidosis, but the effect of fatigue on these processes was stronger in neurosarcoidosis. Neuroimmunological parameters like TNF-α and soluble interleukin-2 receptor serum concentrations do not seem to modulate the pattern of results. Concerning fatigue it seems to be the ‘cognitive’ dimension and not the ‘motor’ dimension that is of relevance for the modulation of response selection in conflicting situations. & 2015 Elsevier B.V. and ECNP. All rights reserved.

n

Corresponding author. Tel.: +49 351 458 7072; fax: +49 351 458 7318. E-mail address: [email protected] (C. Beste). 1 These authors contributed equally. http://dx.doi.org/10.1016/j.euroneuro.2015.01.012 0924-977X/& 2015 Elsevier B.V. and ECNP. All rights reserved.

1.

Introduction

Fatigue is a major symptom in several neuroimmunological diseases (Kluger et al., 2013; Bansal et al., 2012; Greim et al., 2007) including neurosarcoidosis, a rare central nervous system

Fatigue and cognitive control manifestation of sarcoidosis (Lacomis, 2011) (incidence: 2:1,000,000). In neurosarcoidosis, macrophages release cytokines, such as tumor necrosis factor alpha (TNF-α) and interleukins (IL), including IL-2, IL-6, IL-12, IL-15, IL-16 and IL-18 (Lacomis, 2011; Hoitsma et al., 2004). It is likely that fatigue in neurosarcoidosis emerges as a consequence of these immunological alterations, which have often been reported to underlie fatigue (for review: Bansal et al., 2012). Fatigue has been shown to be associated with dysfunctions in cognitive control processes (for review: Boksem and Tops, 2008; Capuron et al., 2005). Lorist and Jolij (2012) showed that the effects of fatigue on the ability to control conflicting information during response selection emerged as a consequence of deficits in top-down attentional processing (Boksem et al., 2006). Similar results have been reported by Van Den Eede et al. (2011) who showed that psychomotor functions are strongly compromised by fatigue. Psychomotor and cognitive control processes are known to be modulated by monoamines and especially the dopamine system (e.g. Willemssen et al., 2009). As the dopamine system has also been implicated in the emergence of fatigue (Lindqvist et al., 2013; Moeller et al., 2012; Meeusen and Roelands, 2010) it seems plausible that cognitive control processes are modulated by fatigue. Despite conclusive results showing an impact of fatigue on cognitive control functions it is unknown in how far fatigue modulates these important cognitive functions in neurosarcoidosis. However, fatigue is not a uni-dimensional construct but is composed of different features, some being more physical and motor-related and some being more cognitive in nature (e.g. Penner et al., 2009). Especially in the context of psychomotor functions and cognitive control this is important to distinguish, since both of these features may be critical for response monitoring functions. Goal of this study is therefore to examine which aspects of fatigue (cognitive or motor aspects) are most important for the modulation of response selection processes and possible deficits in cognitive control in neurosarcoidosis. In the present study we investigate conflict monitoring and response selection processes in neurosarcoidosis patients and healthy controls in relation to cognitive and motor fatigue using neurophysiological techniques. Using event-related potentials (ERPs) conflict monitoring and response selection processes are reflected by the fronto-central N2 component (e.g. Folstein and Van Petten, 2008; Van Veen and Carter, 2002). We hypothesize that neurosarcoidosis patients are less capable to select between two antagonistic response tendencies. This deficit should be expressed by a reduced and/or delayed fronto-central N2 and by increased reaction times when confronted with two antagonistic response tendencies. Interindividual differences in response selection and conflict monitoring may be modulated on the basis of the interindividual degree of fatigue as previous results already report an effect of fatigue on conflict monitoring functions (Lorist and Jolij, 2012). We hypothesize that higher fatigue levels are related to stronger deficits during response selection in neurosarcoidosis patients and controls. However, it is possible that also attentional selection processes are altered and modulate response selection processes. This is because TNF-α, for example, which has been shown to modulate mechanisms of visual attentional selection (Gajewski et al., 2013; Beste et al., 2010), is altered in neurosarcoidosis (Hoitsma et al., 2004) and plays a role in fatigue (e.g. Bansal et al., 2012). To examine the

523 role of potentially altered attentional selection processes in neurosarcoidosis, we examine the visual P1 and N1 that are well-known to reflect visuo-perceptual (P1) and attentional selection processes (N1), respectively (e.g. Herrmann and Knight, 2001). Yet, we do not expect that these functions are altered in neurosarcoidosis and related to interindividual variations in the degree of fatigue. This is because fatigue has been shown to be closely related to the dopaminergic system (for review: Meeusen and Roelands, 2010), which only indirectly modulates bottom-up attentional selection processes and visuo-perceptual processes (Sarter et al., 2006), as measured using the N1 and P1. Due to the neuroimmunological nature of this disease we further examine whether changes in above-mentioned cognitive subprocesses affected in neurosarcoidosis may be modulated by proinflammatory cytokines such as TNF-α and the soluble interleukin-2 receptor (sIL-2R). TNF-α and sIL-2R may be particularly relevant in this regard because these proinflammatory cytokines are assumed to play a major role in the pathogenesis of neurosarcoidosis (Lacomis, 2011; Petereit et al., 2010; Hoitsma et al., 2004) and have been found to be associated with fatigue in other neurological conditions (Lindqvist et al., 2013, 2012; Rudick and Barna, 1990).

2. 2.1.

Experimental procedures Patients and participants

Thirty patients with neurosarcoidosis (NSA) were enrolled into the study. Five patients had to be excluded because of poor quality of the neurophysiological data in the experiment conducted to examine response selection processes. Data analysis was conducted with the remaining N=25 patients. Diagnosis of neurosarcoidosis was done according to the criteria proposed by Zajicek et al. (1999). Zajicek's category ‘possible’ is defined as neurological presentation and the exclusion of possible alternative diagnoses to NSA. The category ‘probable’ also includes proof of a systemic sarcoidosis (by biopsy including the Kveim test and/or two of the following indirect indicators: a Gallium scan, chest imaging and angiotensin-converting enzyme [ACE] in serum) and CNS inflammation (elevated proteins and/or cells, oligoclonal bands and/or a compatible MRI). The category ‘definite’ is established via a biopsy of the nervous system, neurological presentation and the exclusion of alternative diagnoses. We added laboratory parameters such as sIl-2R, TNF-α and b2-microglobulin in serum and cerebrospinal fluid (CSF) to obtain information about the actual inflammation status of the patient (Hoitsma et al., 2004). In the group of patients with neurosarcoidosis, 22 presented a probable diagnosis, and 3 presented a definite diagnosis. Detailed characteristics of the individual neurosarcoidosis patients are shown in Table 1. The neurosarcoidosis patients were recruited from the clinic and outpatient clinic of the Department of Neurology, St. Josef Hospital, RuhrUniversität Bochum. Together with this sample of neurosarcoidosis patients, a sample of case-matched healthy controls (N=25) was recruited by newspaper announcements. The cases were matched for age, sex and educational background. To examine the impact of fatigue on conflict processing and response monitoring functions we used the Fatigue Scale for Motor and Cognitive Functions (FSMC) (Penner et al., 2009). The FSMC is suitable to distinguish between ‘cognitive’ and ‘motor’ fatigue. The FSMC has a high sensitivity, specificity, reliability and validity. In addition to response selection processes we examined basic neuropsychological performance in neurosarcoidosis patients and controls using

524

Table 1 Diagnostic data of the patients enrolled into the study including year of birth and sex. Column 4–10 denote parameters related to the presentation of NS, elevated cell count (44 cells/ml), proteins (4500 mg/l) and oligoclonal bands (OCBs) in cerebrospinal fluid, findings in MRI subdivided on the basis of parenchymal, leptomeningeal and vascular localizations and positive neural or rather meningeal biopsies. Diagnostic evidences for extraneural sarcoidosis such as disease-compatible biopsies of extraneural tissues (e.g. lungs, lymph nodes, skin, and viscera), findings in thoracic and extrathoracic X-ray, computed tomography, bronchoalveolar bavage and elevated SA-dependent laboratory parameters in serum such as ACE, sIL-2R, TNF-α, beta2-microglobulin and neopterin are presented in columns 11–13. No. Year of Sex Neurological birth presentation

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

1964 1967 1941 1966 1966 1945 1986 1955 1991 1975 1976 1976 1962 1986 1982 1961 1981 1966 1977 1962 1961 1956 1976 1952 1983

F F M F F F M M F F F M F F F F F M M F F M M F M

+ + + + + + + + + + + + + + + + + + + + + + + + +

CSF: CSF: elevated cell elevated count proteins + + + + + + + + + + + + + + + + + + +

+ + + + + + + + + + + +

OCB MRI Types of MRIfindings +

+ ND +

+ +

+ + + + +

+ + +

+ +a +a + +a + + +a +a +a + +a +a + +a +a

L (P) P (L) P L,P P V P (L) L,P P (P) L,P L P

+a P + +a +a +a + +

L,P (P) P P P,L P

Neural/ meningeal biopsy

Extra neural biopsy

(Extra-) thoracic Laboratory evidence for extraneural support serum/ SA CSF

ND ND ND ND ND ND ND ND ND ND + ND ND + ND ND ND ND ND ND ND ND ND ND +

+ + +

X/CT X/CT

+ + + ND + + ND + +

+

+ + + +

CT/BAL X CT/BAL X/CT X/CT CT X/CT/BAL (X) Xn X/CT/BAL X CT X/CT/BAL X/CT CT/BAL CT/BAL X/CT/BAL X CT/BAL X

1/2/3/2b 2/3/4b 3/4/4b 2/3 1/2/1b 1/3/1b 1/2/3 2/3/4b 2/2b 2 1/3/1b 3/4 2/3 3/2b/4b 4/4b 2 1b 1/3 1/2/1b 2/3/1b/2b 1/2/4/1b 1/2/3 1/2/3 3/5

AD Diagnostic probability + + + + + + NA + + + + + + + + + + + + + + + + + +

Probable Probable Probable Probable Probable Probable Probable Probable Probable Probable Definite Probable Probable Definite Probable Probable Probable Probable Probable Probable Probable Probable Probable Probable Definite

C. Beste et al.

The abbreviations are as follows: +: pathologies found, : absence of pathologies, ():most likely, MRI: Magnetic Resonance Imaging, OCB: oligoclonal bands, F: female, M: male, L: leptomeningeal, P: parenchymal, V: vascular, SA: sarcoidosis, AD: alternative diagnosis; CSF: cerebrospinal fluid, ND: non-determined; NA: not available, X: X-ray, CT: computed tomography, BAL: bronchoalveolar lavage. n X-ray findings in bones, 1: Angiotensin-Converting Enzyme (ACE), 2: soluble Interleukin-2 receptor (sIL-2R), 3: Tumor Necrosis Factor alpha (TNFα), 4: beta2-microglobulin, 5: Neopterin. a Pathologic MRI, but not further specified. b Elevated laboratory parameter in cerebrospinal fluid.

Fatigue and cognitive control standard neuropsychological tests. These included the trial making test (TMT A/B) to assess general processing speed and the digit span (forward and backward) to assess working memory capacity. Moreover, the Beck depression inventory (BDI-II) was applied to examine the degree of depressive symptoms, which is important to control when examining effects of fatigue on cognitive processes (e.g. Lindqvist et al., 2013; Dantzer et al., 2008). The study was approved by the ethics committee of the Ruhr-University of Bochum. The study was conducted in accordance with the Declaration of Helsinki.

2.2.

Task

To examine response selection processes in NSA a classical flanker task was used (e.g. Beste et al., 2011). It was structured as follows: vertically arranged visual stimuli were presented. The targetstimulus (arrowhead) was presented in the center with the arrowhead pointing to the left or right. The central stimuli were flanked by two vertically adjacent arrowheads which pointed in the same (compatible) or opposite (incompatible) direction as the target. In case of target stimuli (arrowheads pointing to the left or right) participants were required to press a response button with their left or right thumb. The flankers preceded the target by 100 ms. The target (arrowheads or circles) was displayed for 300 ms. The response–stimulus interval was 1600 ms. Flankers and target were switched off simultaneously. Time pressure was administered by asking the subjects to respond within 600 ms. In trials with reaction times exceeding this deadline a feedback stimulus (1000 Hz, 60 dB SPL) was given 1200 ms after the response; this stimulus had to be avoided by the subjects. Four blocks of 105 stimuli each were presented in this task. Compatible (70%) and incompatible stimuli (30%) were presented randomly (c.f. Beste et al., 2011).

2.3.

EEG recording and analysis

The EEG was recorded from 64 Ag/AgCl electrodes located at standard scalp positions according to the extended 10–20 system using a QuickAmp amplifier (Brain Products Inc.). The reference electrode during recording was Cz (sampling rate 1 kHz; electrode impedances o5 kΩ). The EEG data was inspected visually to remove technical artifacts. Afterwards, a band-pass filter ranging from .5 to 20 Hz (48 db/oct) was applied and data inspected for a second time. To correct for periodically recurring artifacts (pulse artifacts, horizontal and vertical eye movements and blinks) an independent component analysis (ICA; Infomax algorithm) was applied to the unepoched data set. Afterwards, the EEG data was segmented according to the different conditions (compatible and incompatible trials). Segmentation was applied with respect to the occurrence of the target stimulus (i.e., stimulus-locked). Automated artifact rejection procedures were applied after epoching: rejection criteria included a maximum voltage step of more than 60 μV/ms, a maximal value difference of 150 μV in a 250 ms interval or activity below .1 μV. Then the data was CSD-transformed (current source density transformation) in order to eliminate the reference potential from the data. A second advantage of the CSD-transformation is that it serves as a spatial filter (Nunez et al., 1991), which makes it possible to identify electrodes that best reflect activity related to cognitive processes. After CSD transformation the baseline correction was performed. For the baseline correction we choose a time window from 800 to 600 ms (i.e., flanker stimulus presentation). The P1, N1 and P3 ERPs were quantified, based on the scalp topography; i.e., electrodes used for data quantification were selected in a data-driven manner. The peak of the respective components was defined as the maximum negativity (N1 and N2) or positivity (P1) within in predefined interval. The intervals were as follows: The visual P1 and N1 were measured at electrode PO7 and PO8 (P1: 0–140 ms; N1: 150–250), the N2 was quantified at

525 electrode FCz (180–400 ms). The ERP components were quantified relative to this pre-flanker baseline. All components were quantified in peak amplitude and latency on the single subject level.

2.4.

Measurement of sIL-2R and TNF-α

Soluble IL-2R and soluble TNF-α were measured in blood serum samples using a solid-phase two-site chemiluminescent immunometric assay via the Siemens IMMULITE 1000 Analyzer (Siemens, Inc., Erlangen, Germany). The coefficient of variation for soluble IL-2R was 3.11 and 3.55 for test and re-test, respectively. The coefficient of variation for soluble TNF-α was 5.99 and 6.33 for test and re-test, respectively. The re-test reliability for sIL-2R and TNF-α was r4.95. The serum samples used for cytokine analysis were drawn  10 min before the patients underwent EEG examination. All patients were tested between 10 a. m. and 12 a.m. after normal breakfast at 7.30 a.m. Concentrations of TNF-α and sIL-2R were integrated with the data on fatigue and cognitive-neurophysiological measures on response control by means of regression analyses.

2.5.

Statistics

The behavioral and electrophysiological data were analyzed using mixed effects ANOVAs. In these ANOVAs, “condition (compatible vs. incompatible)” served as within-subject factor, “group” (neurosarcoidosis vs. control) served as between-subject factor. The factor “electrode” was used as additional within-subject factor for the analysis of the visual P1 and N1. For the electrophysiological data an additional within-subject factor “electrode” was introduced wherever necessary. All variables subjected into the ANOVAs were normally distributed as indicated by Kolmogorov–Smirnov tests (all zo.7; p4.7). Greenhouse–Geisser correction was applied and all post-hoc tests were Bonferroni-corrected.

3. 3.1.

Results Neuropsychological data

For both of the FSMC scores (motor and cognitive score), neurosarcoidosis patients scored higher than controls; cognitive score (neurosarcoidosis: 31.2471.78, controls: 19.287 1.20; t48 =5.55; po.001); motor score (neurosarcoidosis: 33.48 71.72, controls: 1971.25; t48 =6.79; po.001). The ‘cognitive’ and the ‘motor score’ are known to be highly correlated (c.f. Penner et al., 2009), as it was the case for the current sample (r=.766; po.001). For the neuropsychological tests used, there was no difference between neurosarcoidosis patients and controls (all to.5; p4.4). However, in the Beck Depression Inventory, the neurosarcoidosis group showed a higher BDI score (15.3671.67) than the control group (8.12 72.05) (t48 =2.73; p=.004).

3.2.

Behavioral data

The analysis of the reaction times (RTs) revealed a main effect compatibility (F(1,48) = 68.97; po.001; η2 = .59) showing that RTs were longer on incompatible (37979), compared to compatible trials (442711). The main effect “group” further shows that RTs were longer for the NSA group (4347 13), compared to the control group (F(1,48) = 5.05; p= .023; η2 = .103). Importantly, there was an interaction “compatibility  group” (F(1,48) = 17.14; po.001; η2 = .26), which is shown in Figure 1. Post-hoc tests showed that there was no

526 group difference on compatible trials (t48 = .786; p4.2), but on incompatible trials (t48 = 3.33; p= .002). Consequently, the compatibility effect was larger in NSA (91713) than in controls (3177) (t48 = 4.14; po.001). Regarding the error rates, there was a main effect “compatibility” (F(1,48) = 127.37; po.001; η2 = .72) showing that error rates were higher on compatible (3.387.16) than on incompatible trials (7.047.25). The main effect “group” (F (1,48)= 27.91; po.001; η2 = .36) shows that error rates were higher in NSA than in controls. However, since there was no interaction “compatibility  group” (F(1,48) =.97; p4.3), differential effects in compatible and incompatible trials across the groups tested are not due to a speed-accuracy trade-off. The behavioral data suggest for a circumscribed response selection deficit in NSA in incompatible trials. When using the BDI as a covariate to control for possible modulating effects of depressive symptoms the results remain unchanged (all Fo.5; p4.6).

Figure 1 Reaction time (RT) data on compatible and incompatible trials for the neurosarcoidosis patient group and the control group. Error bars denote the standard error of mean.

C. Beste et al.

3.3.

Electrophysiological data

First, the visual P1 and N1 were analyzed. The visual P1 and N1 is shown in Figure 2. As can be seen in Figure 2, the P1 and the N1 ERPs largely overlap in the two groups, and were also similar for compatible and incompatible trials. This pattern is shown in the statistical analysis, where there was no main effect of “group”, “compatibility” or an interaction “compatibility  group” for the P1 and N1 latencies and amplitudes (all Fo1.2; p4.3). The fronto-central N2 is shown in Figure 3. The mixed effects ANOVA revealed a main effect “compatibility” (F(1,48) = 36.60; po.001; η2 = .43) showing that the N2 was smaller (i.e., less negative) on compatible ( .517.46) than on incompatible trials ( 2.037.39).

Figure 3 The fronto-central N2 ERP at electrode FCz. The maps denote the scalp topography plots of the N2 for the different groups. The different colors of the ERP traces denote the different conditions in combination with the different group as denoted in the figure.

Figure 2 The visual P1 and N1 event-related potential (ERP) at electrode PO7 and PO8 for compatible and incompatible trials in the control group and the NSA group. The maps denote the scalp topography plots of the P1 and N1 ERPs for the different groups. The different colors of the ERP traces denote the different conditions in combination with the different group as denoted in the figure.

Fatigue and cognitive control There was also a main effect “group” showing that the N2 was smaller in NSA patients ( .237.22), compared to controls ( 1.997.22) (F(1,48) = 38.35; po.001; η2 = .44). Importantly, there was an interaction “compatibility  group” (F(1,48) = 8.76; p =.005; η2 =.15). Post-hoc tests showed that there was no difference between groups on compatible trials (t48 = .94; p4.2), but group differences were evident on incompatible trials (t48 = 6.68; po.001). On incompatible trials, the N2 was larger in the control group than in NSA patients group (Figure 3). This interaction effect parallels the interaction found for the reaction time data on compatible and incompatible trials. There were generally no main or interaction effects on the latency of the N2 (all Fo1.5; p4.2). As with the behavioral data, there was no effect of the BDI score on the pattern of neurophysiological results (all Fo.8; p4.5).

3.4.

Regression analyses

The above analyses show that NSA patients encounter response selection deficits in incompatible trials (i.e., conflicting situations). This was shown for the behavioral and neurophysiological data. To examine the effect of cognitive and motor fatigue we here perform regression analyses. Reaction times and the N2 on incompatible trials severed as dependent variables in separate regression analyses. The ‘cognitive’ and the ‘motor score’ of the FSMC, as well as the factor group severed as predictor in these models. Using the regression model it is possible to examine the relative influence of even highly inter-correlated predictors, as it is the case for the motor and cognitive fatigue score. For the regression analyses the ‘inclusion method’ was used, i.e., all predictors used were submitted at once in the regression model and not in a stepwise fashion. The scatterplots denoting these correlations for each group are shown in Figure 4. For RTs on incompatible trials (Figure 4A), the regression model was significant (F(2,47) = 53.82; po.001). It is shown that the FSMC ‘cognitive score’ (β = .977; t = 8.70; po.001) and that the factor “group” (β = .228; t= 2.03; p= .045)

527 predicted a significant amount of variance. No effect of the FSMC ‘motor score’ was evident (t = .24; p4.2). As can be seen in Figure 4A the regression line was steeper in the NSA patients than in controls showing that similar increases FSMC ‘cognitive score’ lead to stronger changes in RTs in this group. For N2 on incompatible trials (Figure 4B) also revealed a significant regression model (F(2,47)= 151.58; po.001). This model revealed a significant effect of FSMC ‘cognitive score’ (β = .551; t= 6.67; po.001) and of “group” (β = .493; t = 6.43; po.001). As with the reaction time data, the slope of the regression line was also steeper for the NSA patients group and there was no additional influence of the FSMC ‘motor score’ (t = .19; p4.4). Yet, it is important to control for the effect of depression (e.g. Lindqvist et al., 2013; Dantzer et al., 2008), also since the groups differed in the BDI score. Using multiple regression analyses it is possible to examine and control the effects of intercorrelated regressors (Freedman et al., 2007). To do so we added the BDI score as an additional predictor to the model using the ‘inclusion method’; i.e., the model then included FSMC ‘cognitive score’, ‘group’ and ‘BDI’ as predictors. Using this approach it is possible to control in how far the FSMC ‘cognitive score’ predicts performance and neurophysiological data unbiased of effects related to depression. It is shown that the BDI score did not predict additional variance (β =.122; t =.67; p4.5), while beta weights of the other two regressor (i.e., FMSC ‘cognitive score’ and ‘group’) remained still significant (po.001). This shows that the correlations obtained for the FSMC score are robust against the influence of depressive symptoms. Using blood serum concentrations of sIl-2R, TNF-α as additional predictor in the model did also not explain more variance in the dependent variable (all β = .199; t = .99; p4.3). There were also no correlations between the degree of cognitive and motor fatigue and blood serum concentrations of sIl-2R, TNF-α (all ro.072; p4.3). Opposed to data reflecting response selection, there were generally no significant correlations between the degree of fatigue on either the ‘cognitive’ or the ‘motor’ score concerning processing speed measured using the TMT (ro.066; p4.2),

Figure. 4 (A) Scatterplot denoting the correlation between the reaction times on incompatible trials and the FSMC cognitive score for the NSA group (black dots) and controls (white dots). Note that the slopes of the regression lines differ in their steepness. (B) Scatterplot denoting the correlation between the N2 amplitude on incompatible trials and the FSMC cognitive score for the NSA group (black dots) and controls (white dots).

528

C. Beste et al.

or working memory capacity (ro.112; p4.3). Similarly, there were no correlations of fatigue with the interindividual modulation of the P1 and N1 ERP (ro.09; p4.4).

4.

Discussion

In the current study we examined how response selection and conflict monitoring processes are modulated in a rare neuroimmonological disease (neurosarcoidosis) and how fatigue as a major symptom in this rare disease modulates interindividual differences in these executive control processes. The results show that reaction times differed between controls and neurosarcoidosis patients only in the incompatible condition. In the incompatible condition reaction times were slower in neurosarcoidosis than in controls. Together with the finding that there was no difference in reaction times on compatible trials, the results suggest that there was a circumscribed deficit to select between different conflicting responses in neurosarcoidosis patients. The neurophysiological data provide insights into the mechanisms that may mediate difficulties in response selection in neurosarcoidosis. The results show that there was no difference between neurosarcoidosis patients and controls in the visual P1 and N1 ERP. In line with other studies (Beste et al., 2014), this shows that visuo-perceptual processes and attentional selection processes (as reflected by the visual P1 and N1) (e.g. Herrmann and Knight, 2001) do not underlie deficits observed in neurosarcoidosis and suggest that these cognitive functions are spared in neurosarcoidosis. Opposed to attentional selection processes, the fronto-central N2 was smaller in neurosarcoidosis patients, compared to controls. Several interpretations of the N2 have been put forward in the last years, namely that the N2 reflects conflict monitoring (e.g. Botvinick et al., 2004) or response selection processes (e.g. Gajewski et al., 2008) (for review: Folstein and Van Petten, 2008). For the following reasons, the current results would speak for deficits in response selection rather conflict monitoring functions: Reaction times were higher in neurosarcoidosis in the incompatible condition. If deficits in conflict monitoring functions underlie these effects, an increased N2 in neurosarcoidosis should have been evident. Yet, the opposite was the case. As such the results suggest that deficient processes related to the selection between different responses underlie the deficits observed in neurosarcoidosis patients. These findings, that action control processes are specifically altered, fit well to other studies showing that the selection of actions is dysfunctional under fatigue (Boksem et al., 2006; Lorist et al., 2000). Interestingly, response selection processes as reflected by the N2 are known to be mediated by the dopaminergic system (e.g. Willemssen et al., 2009, 2011) and this system is known to mediate fatigue effects (e.g. Moeller et al., 2012; Meeusen and Roelands, 2010). It may be a question of future research to examine if there are alterations in dopaminergic functions in neurosarcoidosis. Concerning the precise interrelation of fatigue and response selection processes the regression analyses revealed that interindividual differences in responses selection, on the behavioral and reaction time level, are predictable by the interindividual varying degree of cognitive fatigue. Of note, no effect of motor fatigue was evident in controls and

neurosarcoidosis and the level of depression did not further modulate the results. This suggests that it is the cognitive dimension of fatigue that modulates response selection processes. The effect of ‘cognitive fatigue’ seems to be highly specific for response selection processes, since there was no correlation with classical neuropsychological tests assessing processing speed and working memory. In this regard, and in line with other studies (e.g. Lorist and Jolij, 2012; Van Den Eede et al., 2011; for review: Boksem and Tops, 2008) fatigue predominantly seems to affect cognitive control processes, but seems to leave other cognitive functions not closely related to cognitive control unaffected. With respect to the influence of fatigue on cognitive control (response selection processes), the regression analyses further show that the regression line was steeper showing that similar increases in fatigue lead to greater effects on responses selection as it was the case for the control group. Hence, similar increases in fatigue have a stronger effect on response selection processes in neurosarcoidosis patients than in controls. It is possible that this is an effect of the general fatigue level, which was higher in neurosarcoidosis patients. Once a certain degree of fatigue is reached, further increases lead to stronger effects on cognitive functions. Of note, there were no correlations with examined TNF-α and sIL-2R serum concentrations, which suggest that at least these neuroimmunological parameters do not mediate the effects observed. Previous studies provided evidence for a role of sIL-2R serum concentrations for response control processes when action cascading processes (i.e., stopping and changing of responses) were examined (Beste et al., 2014). However response selection processes examined in this study that are reflected by the N2 have not been shown to be modulated in action cascading (e.g. Mückschel et al., 2014; Stock et al., 2014; Yildiz et al., 2014). The processes examined in the current study are different to previous study by Beste et al. (2014). This likely account for the between-study differences in the relevance of sIL-2R serum concentrations for cognitive control processes. Moreover, in previous studies (e.g. Lindqvist et al., 2012) correlations with fatigue were only small for sIL-2R and absent for TNF-α. It may be argued that the lack of correlation reflects a lack of power due to the relatively small sample size. However, the fact in other studies in NSA (Beste et al., 2014) with similar sample sizes reliable correlations were obtained makes this reason unlikely. The current results do not exclude the possibility of association between CNS or CSF cytokines and cognitive control. In addition, the study does not exclude the association of other cytokines such as IL6, IL-12, IL-15, IL-16 and IL-18. An examination of the factors may be subject to future research using more extended sample sizes. In summary, the study shows that neurosarcoidosis is associated with cognitive deficits related to the selection between different responses. No changes in attentional selection processes were evident. Neurosarcoidosis patients do not show a conflict monitoring deficit, but dysfunctions in selecting the appropriate motor program. Of note, fatigue strongly modulates responses selection processes in controls and neurosarcoidosis, but the effect of fatigue on response selection processes was stronger in neurosarcoidosis, compared to controls. Concerning fatigue it seems to be the ‘cognitive’ dimension that is of relevance for the modulation of response selection processes.

Fatigue and cognitive control

Role of the funding source The funding source played no role in the design of study, data collection and analyses or discussion of the results, or in the decision to submit and publish the data.

Contributors Janina Kneiphof performed the experiments and wrote the paper. Dirk Woitalla conceived the study and provided materials and wrote the paper. Christian Beste conceived the study, analyzed the data and wrote the paper.

Conflicts of interest There are no conflicts of interest.

Acknowledgments This work was supported by a Grant from the Deutsche Forschungsgemeinschaft (DFG) BE4045/10-1 and 10-2 to C.B.

References Bansal, A.S., Bradley, A.S., Bishop, K.N., Kiani-Alikhan, S., Ford, B., 2012. Chronic fatigue syndrome, the immune system and viral infection. Brain Behav. Immun. 26, 24–31. Beste, C., Kneiphof, J., Woitalla, D., 2014. Modulatory effects of proinflammatory cytokines on action cascading processes – evidence from neurosarcoidosis. Brain Behav. Immun. 41, 126–133. Beste, C., Güntürkün, O., Baune, B.T., Domschke, K., Falkenstein, M., Konrad, C., 2011. Double dissociated effects of the functional TNFalpha-308G/A polymorphism on processes of cognitive control. Neuropsychologia 49, 196–202. Beste, C., Baune, B.T., Falkenstein, M., Konrad, C., 2010. Variations in the TNF-alpha gene (TNF-alpha-308G-A) affect attention and action selection mechanisms in a dissociated fashion 104, 2523–2531. Boksem, M.A., Tops, M., 2008. Mental fatigue: costs and benefits. Brain Res. Rev. 59, 125–139. Boksem, M.A., Meijman, T.F., Lorist, M.M., 2006. Mental fatigue, motivation and action monitoring. Biol. Psychol. 72, 123–132. Botvinick, M.M., Cohen, J.D., Carter, C.S., 2004. Conflict monitoring and anterior cingulate cortex: an update. Trends Cogn. Sci. 8, 539–546. Capuron, L., Pagnoni, G., Demetrashvili, M., Wollwine, B.J., Nemeroff, C.B., Berns, G.S., Miller, A.H., 2005. Anterior cingulate activation and error processing during interferon-alpha treatment. Biol. Psychiatry 58, 190–196. Dantzer, R., O’Connor, J.C., Freund, G.G., Johnson, R.W., Kelley, K.W., 2008. From inflammation to sickness and depression: when the immune system subjugates the brain. Nat. Rev. Neurosci. 9, 46–56. Folstein, J.R., Van Petten, C., 2008. Influence of cognitive control and mismatch on the N2 component of the ERP: a review. Psychophysiology 45, 152–170. Freedman, D., Pisani, R., Pruves, R., 2007. Statistics, fourth ed. Norton & Company, New York. Greim, B., Engel, C., Apel, A., Zettl, U.K., 2007. Fatigue in neuroimmunological diseases. J. Neurol. 254 (Suppl. 2), S102–S106. Gajewski, P.D., Hengstler, J.G., Golka, K., Falkenstein, M., Beste, C., 2013. The functional tumor necrosis factor-alpha (308G/A)

529 polymorphism modulates attentional selection in elderly individuals. Neurobiol. Aging 34, 2694.e1–2694.e12. Gajewski, P.D., Stoerig, P., Falkenstein, M., 2008. ERP-correlates of response selection in a response conflict paradigm. Brain Res. 1189, 127–134. Herrmann, C.S., Knight, R.T., 2001. Mechanisms of human attention: event-related potentials and oscillations. Neurosci. Biobehav. Rev. 25, 465–476. Hoitsma, E., Faber, C.G., Drent, M., Sharma, O.P., 2004. Neurosarcoidosis: a clinical dilemma. Lancet Neurol. 3, 397–407. Kluger, B.M., Krupp, L.B., Enoka, R.M., 2013. Fatigue and fatigability in neurologic illnesses: proposal for a unified taxonomy. Neurology 80, 409–416. Lacomis, C., 2011. Neurosarcoidosis. Curr. Neuropharmacol. 9, 429–436. Lindqvist, D., Hall, S., Surova, Y., Nielsen, H.M., Janelidze, S., Brundin, L., Hansson, O., 2013. Cerebrospinal fluid inflammatory markers in Parkinson's disease—associations with depression, fatigue and cognitive impairment. Brain Behav. Immun. 33, 183–189. Lindqvist, D., Kaufman, E., Brundin, L., Hall, S., Surova, Y., Hansson, O., 2012. Non-motor symptoms in patients with Parkinson's disease – correlations with inflammatory cytokines in serum. Plos One 7, e47387. Lorist, M.M., Jolij, J., 2012. Trial history effects in stroop task performance are independent of top-down control. Plos One 7, e39802. Lorist, M.M., Klein, M., Nieuwenhuis, S., De Jong, R., Mulder, G., Meijman, T.F., 2000. Mental fatigue and task control: planning and preparation. Psychophysiology 37, 614–625. Meeusen, R., Roelands, B., 2010. Central fatigue and neurotransmitters, can thermoregulation be manipulated? Scand. J. Med. Sci. Sports Suppl. 3, S19–S28. Moeller, S.J., Tomasi, D., Honorio, J., Volkow, N.D., Golstein, R.Z., 2012. Dopaminergic involvement during mental fatigue in health and cocaine addiction. Transl. Psychiatry 2, e176. Mückschel, M., Stock, A.K., Beste, C., 2014. Psychophysiological mechanisms of interindividual differences in goal activation modes during action cascading. Cereb Cortex 24, 2120–2129. Nunez, P.L., Pilgreen, K.L., Westdorp, A.F., Law, S.K., Nelson, A.V., 1991. A visual study of surface potentials and Laplacians due to distributed neocortical sources: computer simulations and evoked potentials. Brain Topogr. 4, 151–168. Penner, I.K., Raselli, C., Stöcklin, M., Opwis, K., Kappos, L., Calabrese, P., 2009. The fatigue scale for motor and cognitive functions (FSMC): validation of a new instrument to assess multiple sclerosis-related fatigue. Mult. Scler. 15, 1509–1517. Petereit, H.F., Reske, D., Tumani, H., Jarius, S., Markus Leweke, F., Woitalla, D., Pfister, H.W., Rubbert, A., 2010. Soluble CSF interleukin 2 receptor as indicator of neurosarcoidosis. J. Neurol. 257, 1855–1863. Rudick, R.A., Barna, B.P., 1990. Serum interleukin 2 and soluble interleukin 2 receptor in patients with multiple sclerosis who are experiencing severe fatigue. Arch. Neurol. 47, 254–255. Sarter, M., Gehring, W.J., Kozak, R., 2006. More attention must be paid: the neurobiology of attentional effort. Brain Res. Rev. 51, 145–160. Stock, A.K., Arning, L., Epplen, J.T., Beste, C., 2014. DRD1 and DRD2 genotypes modulate processing modes of goal activation processes during action cascading. J. Neurosci. 34, 5335–5341. Van Den Eede, F., Moorkens, G., Hulstijn, W., Maas, Y., Schrijvers, D., Stevens, S.R., Cosyns, P., Claes, S.J., Sabbe, B.G., 2011. Psychomotor function and response inhibition in chronic fatigue syndrome. Psychiatry Res. 186, 367–372. Van Veen, V., Carter, C.S., 2002. The anterior cingulate as a conflict monitor: fMRI and ERP studies. Physiol. Behav. 77, 477–482. Willemssen, R., Falkenstein, M., Schwarz, M., Müller, T., Beste, C., 2011. Effects of aging, Parkinson's disease, and dopaminergic

530 medication on response selection and control. Neurobiol. Aging 32, 327–335. Willemssen, R., Müller, T., Schwarz, M., Falkenstein, M., Beste, C., 2009. Response monitoring in de novo patients with Parkinson's disease. Plos One 4, e4898. Yildiz, A., Quetscher, C., Dharmadhikari, S., Chmielewski, W., Glaubitz, B., Schmidt- Wilcke, T., Edden, R., Dydak, U., Beste, C., 2014. Feeling safe in the plane: neural mechanisms

C. Beste et al. underlying superior action control in airplane pilot trainees – a combined EEG/MRS study. Hum. Brain Mapp. 35, 5040–5051. Zajicek, J.P., Scolding, N.J., Foster, O., Rovaris, M., Evanson, J., Moseley, I.F., Scadding, J.W., Thompson, E.J., Chamoun, V., Miller, D.H., McDonald, W.I., Mitchell, D., 1999. Central nervous system sarcoidosis – diagnosis and management. QJM 92, 103–117.