You are here
A study of cognitive fatigue in Multiple Sclerosis with novel clinical and electrophysiological parameters utilizing the event related potential P300
Multiple Sclerosis and Related Disorders, Volume 10, November 2016, Pages 1 - 6
Although cognitive fatigue plays a significant part in Multiple Sclerosis (MS) related impairment, knowledge regarding it is largely lacking. Until now, not many tools are available to a clinician to detect cognitive fatigue. The subjective tools of fatigue have never been reliable.tabl
To assess the prevalence and clinical/ demographic profile of cognitive fatigue in MS using novel clinical and electrophysiological measures and to find their accuracy. We also aimed to test the three leading hypotheses - the temporal fatigue, cognitive load and cognitive domain hypotheses of cognitive fatigue in MS.
50 consecutive MS patients attending the Neurology OPD in Madras Medical College, Chennai from May 2015 to February 2016 satisfying the 2010 revised McDonald criteria for MS with an equal number of matched controls were recruited. Modified versions (a shorter version, and longer and more demanding versions) of the Stroop test, symbol digit modalities test, and serial addition tests were used in addition to modified tests of P300 latency and amplitude each specifically tailored to reveal cognitive fatigue.
Out of the seven measures of cognitive fatigue used, 46% (n=23) of MS patients had impairment in two or more of the scores compared to that of 8% (n=4) in the healthy control group. The Expanded disability status scale (EDSS) scores were significantly higher for MS patients with cognitive fatigue compared to those without. All the clinical and electrophysiological measures used in this study had a relatively high sensitivity and specificity. In addition, all the clinical measures correlated with the electrophysiological measures of cognitive fatigue in this study. Our data also supported all three hypotheses implying that cognitive fatigue in MS may be a multifaceted entity.
Cognitive fatigue is widely prevalent in MS and can be detected with specific tools. The tools developed and described in this study may be used as an effective means of detecting cognitive fatigue in MS patients and thus allowing patients to realise their limitations. Institution of appropriate remedial measures like advising such patients to break up a more cognitively demanding task into smaller subtasks may help to improve their quality of life.
- The prevalence of cognitive fatigue in Multiple Sclerosis is 46%.
- Novel tools used in this study were effective in revealing cognitive fatigue.
- Novel electrophysiological tools using P300 were also effective.
- Evidence for all three competing hypotheses for cognitive fatigue was obtained.
Keywords: Cognitive fatigue, Multiple Sclerosis, P300 - event related potential.
Fatigue – both physical and cognitive plays a significant part in Multiple Sclerosis (MS) symptomatology. Prevalence estimates of fatigue are as high as 70–90% (Minden et al, 2006, Krupp et al, 1988, Fisk et al, 1994, and Freal et al, 1984). Cognitive fatigue may be objective or subjective and may be the result of central or peripheral processes (Kluger et al, 2013 and Chaudhuri and Behan, 2000). Insights regarding the factors governing cognitive fatigue are lacking and hence so are the remedial counter-measures to tackle them. This study aims to uncover the prevalence of cognitive fatigue in MS and attempts to reveal the value of various novel specifically tailored clinical and electrophysiological measures in its diagnosis.
Various attributes of a cognitive task correlate with the degree of cognitive fatigue for that task. The time required to complete a task seems to be the most important in this regard, which appears to be directly related to the degree of cognitive fatigue (Krupp and Elkins, 2000) and this forms the basis of the temporal fatigue hypothesis. But some studies indicate the opposite – as the time spent on a task increases – performance improves (Ackerman and Kanfer, 2009 and Jensen et al, 2013). This delicate relationship may be due to the opposing influences of fatigue and learning on task performance. Further clarifications are required to comprehend whether impaired learning ability which also requires white matter integrity in addition to fatigue per se is responsible for impaired cognitive performance in MS.
The other competing hypothesis – the cognitive load hypothesis suggests that the degree of cognitive fatigue is proportional to the cognitive load of a task (Bailey et al., 2007). In fact, both these hypotheses may be correct in a given case with fatigue. Finally, the third popular hypotheses in cognitive fatigue is the cognitive domain hypothesis which states that fatigue may be modality restricted (Berrigan et al, 2013, Genova et al, 2009, and DeLuca et al, 2004). However, it might not be pragmatic to tease out the affected domain(s) since no test (or battery of tests) could exclusively probe into a single domain.
The primary objective was to enquire the prevalence and clinical/ demographic profile of cognitive fatigue in MS using novel clinical and electrophysiological measures designed specifically to reveal cognitive fatigue. We also tried to measure sensitivity and specificity of these individual tests in detecting fatigue. Finally, we attempted to check the above-mentioned hypotheses of cognitive fatigue.
50 consecutive MS patients attending the Neurology OPD in Madras Medical College, Chennai from May 2015 to February 2016 satisfying the 2010 revised McDonald criteria (Polman et al., 2011) for MS with an equal number of age, sex, and number of years of education (and level of education) matched controls were recruited. Institutional Ethics Committee approval was obtained beforehand. After an informed consent from the subjects, a detailed history was recorded and a thorough neurological assessment was made. Routine laboratory investigations like blood sugar levels, haemogram, electrolytes, renal and liver function tests, HIV, and VDRL were done to rule out comorbidities. Physical disability was assessed with the EDSS (Expanded Disability Status Scale) (Kurtzke, 1983) and depression was assessed with the CESD (Centre for Epidemiologic Studies Depression) scale (Radloff, 1977). Subjects with a developmental delay, comorbidities like stroke, major psychiatric illnesses, substance abuse, history of brain trauma, impaired colour vision, and a visual acuity of less than 20/100 in either eye were excluded. In addition, all subjects underwent the following additional tests aimed at cognitive fatigue.
In each of the tests mentioned below, the administrator of the test had an answer key for convenience and used a stop watch for timing. Also, the subjects were given three prior trials with a different set of ‘mock’ test items to get them acquainted to these tests. A sample of each of these tests mentioned below is provided in Fig. 1. Subjects were also given a mandatory resting period of 20 min between tests to minimise the effect of cognitive fatigue on the successive test scores. All the patients were ensured to have had at least 8 h of sleep the previous night. Anyone who came with deprived sleep were given appointments on another day to prevent any sleep deprivation related fatigue from contaminating our work.
Sample test items and answer keys used.
A modified version of the Stroop test which is a test of sustained attention and information processing speed was used. We developed a version where the subjects were given a sheet of paper containing 250 words (with font colours incoherent to their semantic meaning) and were instructed to name the colour of the font of the words in the list successively as quickly as possible. The Stroop 60 score was calculated as the number of correct responses in 60 s and the Stroop 180 score was the number of correct responses in 180 s. Incorrect responses were ignored. The number of correct responses was taken as the Stroop 60 and 180 scores respectively. We hypothesised that the (3*Stroop 60) to Stroop 180 ratio would be a measure of cognitive fatigue since it tests and quantitates sustainability of attention and processing speed for a prolonged time period.
A modified version of the symbol digit modalities test (mSDMT) which measures concentration and sustained attention was used. In this, the subjects were presented with a unique code sheet where different symbols were equated to each of the 9 single digit numbers (from 1 to 9) and were also given another test sheet with a sequence of 150 symbols. Then, they were instructed to tell the correct number corresponding to the symbols in sequence as quickly as possible with the help of the code sheet. Wrong answers were ignored. The mSDMT 60 score was the number of correct responses in 60 s and mSDMT 180 was the number of correct responses in 180 s (3*mSDMT 60) to mSDMT 180 ratio was considered a measure of cognitive fatigue.
To increase the cognitive load in the mSDMT test, we developed and used a more cognitively demanding version – the SDMCAT test (Symbol Digit Modality Cum Addition test) were the subjects were presented with a similar code sheet and test sheet, but instead of telling out the correct number represented by the symbols, the subjects were asked to tell out the sum of the numbers represented by the consecutive pair of symbols. Similar to mSDMT 60 and mSDMT 180 scores, SDMCAT 60 and SDMCAT 180 scores were computed and (3*SDMCAT 60) to SDMCAT 180 ratio was considered a measure of cognitive fatigue.
We also used a serial addition test (SAT) for assessing attention, calculation, and information processing speed where the subjects were presented with a sheet of paper containing a random sequence of 250 single digit numbers and were instructed to tell the sum of consecutive pairs as quickly as possible. Wrong answers were ignored and the number of correct responses in 60 s was the SAT 60 score and in 180 s was the SAT 180 score. (3*SAT 60) to SAT 180 ratio was used as a measure of cognitive fatigue.
To measure the effect of increasing cognitive load on fatigue, 3DSAT test (Three digit serial addition test) in which subjects were requested to tell the sum of three consecutive digits in contrast to two digits in the SAT test, was used. Similar to the SAT test the 3DSAT 60 and 3DSAT 180 scores were measured and (3*3DSAT 60) to 3DSAT 180 ratio was used as a measure of cognitive fatigue.
In addition, the evoked potential P300 was measured with the study parameters in Table 1. The subjects were required to count the rare stimuli. We checked for two measures – the P300 peak latency (P300L) and the peak amplitude (P300A). We initially measured P300 latency and amplitude (P300L50 and P300A50 scores respectively) as the average of 50 rare stimuli or 250 frequent stimuli (probability of rare stimuli was set at 0.2). Even after these preliminary measurements were made, the test was continued for a total of 150 rare stimuli or 750 frequent stimuli and latency and amplitude at the end was scored as P300L150 and P300A150 respectively. We hypothesised that P300L150: P300L50 ratio and P300A50: P300A150 ratio would be novel electrophysiological markers for cognitive fatigue since latency increases while amplitude decreases as attention becomes ill sustainable with fatigue.
Study parameters for P300 evoked potential.
|Test name||Auditory evoked P300 - using oddball paradigm|
|Stimulus type||Tone burst|
|Probability of rare stimulus||0.2|
|Tone intensity||75 dB|
|Tone frequency||2000 Hz (rare) and 1000 Hz (frequent)|
|Stimulus speed||1 per second|
|Electrode position||Cz - active, A1 reference|
|Impedance||≤2 KΩ between electrodes|
|Band pass filter||1–25 Hz|
|Randomness of stimuli||Pseudorandom; Since there was a constraint that no two rare stimuli can occur together.|
All measures of cognitive fatigue were tailored in such a way that an increase in any measure indicated an increased cognitive fatigue. From the above mentioned tests we obtained the following 5 clinical measures of cognitive fatigue - (3*Stroop 60) to Stroop 180 ratio, (3*mSDMT 60) to mSDMT 180 ratio, (3*SAT 60) to SAT 180 ratio, (3*SDMCAT 60) to SDMCAT 180 ratio, and (3*3DSAT 60) to 3DSAT 180 ratio and two electrophysiological measures - P300L150: P300L50 ratio and P300A50: P300A150 ratio. Out of the clinical measures, the last two were specifically aimed to test the cognitive load hypothesis while all tests were used to test the temporal fatigue hypothesis.
Subjective measures of cognitive fatigue were not used in this test since several studies published previously have already established no or minimal correlation between the subjective and objective measures of fatigue (Paul et al, 1998, Schwartz et al, 1996, and Parmenter et al, 2003).
SPSS software was used for statistical analyses. Student's T-test and chi square tests for equality of proportions were used to check for significant differences in the demographic variables between the case and control group as well as between patients with significant cognitive fatigue and those without. Student's T-test was also used to check for difference in the mean scores of various cognitive measures and measures of cognitive fatigue between the case and control groups. ROC curves were plotted for all the measures of cognitive fatigue to obtain their sensitivity and specificity as a diagnostic test for cognitive fatigue. Pearson correlation coefficients were calculated between the electrophysiological measures and the clinical measures of cognitive fatigue.
50 cases with various forms of MS and 50 age, sex, educational and employment status matched controls were included. The demographic profile is illustrated in Table 2. The data confirms that there were no significant differences in the sex, age, educational or employment status between the groups. Among 50 patients, 36 (72%) had Relapsing remitting MS. 2 (4%) and 12 (24%) had primary progressive MS and secondary progressive MS respectively. The mean duration of disease was 6.0±7.4, ranging from 0.5 to 50 years.
Subject demographic and disease characteristics, Mean±SD (range).
|Females, n (%)||35 (70%)||34 (68%)||0.8288|
|Age, mean±SD (Range)||33.6 ± 10.6 (13–66)||33.6 ± 9.6 (16–50)||0.9842|
|Education in years, mean±SD (range)||9.4 ± 4.0 (0–18)||10.6 ± 4.2 (0–18)||0.1708|
|Employed, n (%)||14 (28%)||16 (32%)||0.6625|
|RRMS, n (%)||36 (72%)||NA||NA|
|SPMS, n (%)||12 (24%)||NA||NA|
|PPMS, n (%)||2 (4%)||NA||NA|
|Disease duration in years, mean±SD (range)||6.0 ± 7.4 (0.5–50.0)||NA||NA|
|EDSS, mean±SD (range)||4.6 ± 1.9 (1.0–9.0)||NA||NA|
|CESD, mean±SD (range)||11.3 ± 8.7 (0–31)||3.1 ± 3.2 (0–12)||< 0.0001|
EDSS was 4.6±1.9, ranging from 1.0 to 9.0. MS patients had a mean CESD score of 11.3±8.7 ranging from of 0-31, and the controls had a score of 3.1±3.2 ranging from 0 to 12. Student t-test was used for comparing the means. The difference in CESD scores was statistically significant at 95% confidence interval between the populations meaning that MS patients are more depressed compared to their healthy counterparts. None of the patients were on steroids or any other drug which might affect their cognitive performance at the time of testing.
The scores of MS patients and the healthy control group in all the tests are illustrated in Table 3 and the scores in five clinical and 2 electrophysiological measures of cognitive fatigue are represented in Table 4. We can clearly see a pattern from Table 3 – though the Stroop60, mSDMT 60, SDMCAT60, 3DSAT60 scores are not significantly different between the groups, the 180 s version of each of these tests gave a statistically significant difference between the two groups, with the MS group getting poorer scores relatively. This by itself can be viewed as evidence in favour of the temporal fatigue hypothesis of cognitive fatigue in MS. Nonetheless, both SAT60 and SAT180 scores were not significantly different between the groups. Similarly, P300L50 and P300L150 scores were significantly different between the groups, with MS patients having a more prolonged latency. But there were no significant differences between the P300A50 and P300A150 scores.
Performance parameters on various clinical and electrophysiological measures, Mean±SD (range).
|Stroop60||49.62±8.94 (31–65)||51.68±4.28 (46–62)||0.1448|
|Stroop180||145.56±25.09 (84–186)||157.20±14.85 (122–192)||0.0058|
|mSDMT60||31.78±5.83 (22–42)||33.28±3.65 (26–42)||0.1263|
|mSDMT180||96.64±7.75 (57–131)||103.56±11.84 (80–126)||0.0240|
|SAT60||36.10±10.23 (18–54)||37.32±8.11 (22–52)||0.5103|
|SAT180||110.44±33.31 (48–171)||118.64±26.12 (64–164)||0.1739|
|SDMCAT60||8.02±2.74 (1–13)||8.06±2.48 (2–16)||0.9392|
|SDMCAT180||21.78±9.04 (2–39)||26.96±9.99 (7–57)||0.0078|
|3DSAT60||16.10±5.71 (8–28)||17.16±2.83 (12–24)||0.2425|
|3DSAT180||45.48±18.36 (17–86)||56.40±12.24 (40–88)||0.0007|
|P300L50||322.348±4.351 (315.1–331.3)||319.242±5.282 (309.6–331.2)||0.0018|
|P300L150||327.914±6.490 (315.1–340.2)||322.648±6.096 (310.2–339.1)||<0.0001|
|P300A50||6.586±1.099 (4.2–8.52)||6.760±1.530 (3.1–9.8)||0.5141|
|P300A150||6.026±1.103 (3.36–8.4)||6.354±1.453 (2.5–8.8)||0.2072|
Performance parameters on various measures of cognitive fatigue, Mean±SD (range).
|(3*Stroop 60)/Stroop 180||1.02304±0.06266 (0.84615–1.17857)||0.98898±0.05739 (0.89286–1.13115)||0.0056|
|(3*mSDMT 60)/mSDMT 180||0.99102±0.09204 (0.75000–1.31579)||0.96642±0.06058 (0.83607–1.08000)||0.1176|
|(3*SAT 60)/SAT 180||0.99450±0.14624 (0.75824–1.51648)||0.94799±0.08236 (0.78947–1.14000)||0.0529|
|(3*SDMCAT 60)/SDMCAT 180||1.16938±0.22818 (0.90000–1.90909)||0.92104±0.12480 (0.42105–1.09091)||<0.0001|
|(3*3DSAT 60)/3DSAT 180||1.10025±0.18033 (0.78000–1.60000)||0.93162±0.14567 (0.65625–1.28571)||<0.0001|
|P300L150: P300L50||1.01729±0.01663 (0.99969–1.06137)||1.01672±0.00959 (0.99670–1.03542)||0.0165|
|P300A50: P300A150||1.10009±0.09673 (0.97619–1.30769)||1.06716±0.06509 (1.00000–1.28947)||0.0486|
More importantly, from Table 4, the clinical measures of cognitive fatigue, (3*Stroop 60)/Stroop 180, (3*SDMCAT 60)/SDMCAT 180, and (3*3DSAT 60)/3DSAT 180 and both the electrophysiological measures of cognitive fatigue - P300L150: P300L50 and P300A50: P300A150 were significantly different between the groups implying that MS patients did in fact suffer from more cognitive fatigue compared to the healthy controls. However, (3*mSDMT 60)/mSDMT 180 and (3*SAT 60)/ SAT 180 which were also measures of cognitive fatigue did not differ significantly between the groups implying that fatigue may in fact be domain specific giving some support to the cognitive domain hypothesis of fatigue.
We compared the mean scores in (3*mSDMT 60)/mSDMT 180 and (3*SDMCAT 60)/SDMCAT 180 in the MS group using Student's T-test and found a statistically significant difference (p <0.0001). Similar significant difference (p =0.0017) was obtained on comparing the (3*SAT 60)/ SAT 180 scores with the (3*3DSAT 60)/3DSAT 180. These two results indicate that as the load of the cognitive task increases, MS patients become increasingly less efficient in performing them irrespective of the time spent on the task. When a similar comparison was made in the healthy control group, no such difference could be found. In fact, in the control group, the opposite was seen; the (3*SDMCAT 60)/SDMCAT 180 and (3*3DSAT 60)/3DSAT 180 scores were lesser than their less cognitively demanding counterparts - (3*mSDMT 60)/mSDMT 180 and (3*SAT 60)/ SAT 180 respectively. This could be due to the fact that as the load of a cognitive task increases, learning plays a significant role in increasing the performance efficiency in normal subjects which is not the case in MS patients in whom cognitive fatigue nullifies this learning induced performance enhancement. This favours the argument that the cognitive load hypothesis is also valid in explaining the mechanism of cognitive fatigue in MS.
Also, both the electrophysiological measures of cognitive fatigue, the P300L150: P300L50 and P300A50: P300A150 scores were also significantly different between the groups (Table 4).
For the 7 measures of cognitive fatigue described above, we defined an abnormal score as 1.5 standard deviations above the mean value of the healthy controls. The number of individuals with abnormal scores among the 7 different measures of fatigue (with abnormal scores taken as 1.5, 2, 3 SD from mean normal scores) is illustrated in Fig. 2. We also arbitrarily defined MS patients with significant cognitive fatigue as those who had abnormal scores in 2 or more of the 7 measures. 46% (n=23) in the MS group and 8% (n=4) in the healthy control group had a significant cognitive fatigue according to this definition. The rest (n=27) were grouped under ‘MS patients without significant cognitive fatigue’ for the purpose of further comparison. On comparing the groups with and without significant cognitive fatigue (Table 5) we could find no significant difference in gender, age, number of years of education (a surrogate marker for cognitive reserve), employment status, type of MS, disease duration or the CESD scores. However, the EDSS scores were significantly higher in patients with significant cognitive fatigue compared to those without. There was no significant difference in treatment history between these groups.
Number of cases and controls with impaired 0−7 measures of cognitive fatigue at 1.5, 2 and 3 sd from mean normal scores.
Subject demographic and disease characteristics of ms patients with and without significant cognitive fatigue, Mean±SD (range).
|Characteristics||MS patients with significant cognitive fatigue||MS patients without significant cognitive fatigue||p|
|Females, n (%)||16 (69.57%)||19 (70.37%)||0.9522|
|Age, mean±SD (range)||34.5±9.9 (17–66)||32.8±11.2 (13–60)||0.9842|
|Education in years, mean±SD (range)||9.6±4.2 (2–18)||9.3±3.9 (0–16)||0.8160|
|Employed, n (%)||7 (30.43%)||7 (25.93%)||0.7263|
|RRMS, n (%)||16 (69.57%)||20 (74.07%)||0.7263|
|SPMS, n (%)||5 (21.74%)||7 (25.93%)||0.1188|
|PPMS, n (%)||2 (8.70%)||0 (0%)||0.7263|
|Disease duration in years, mean±SD (range)||7.3±9.9 (0.75–50.0)||4.9±4.1 (0.5–15)||0.2750|
|EDSS, mean±SD (range)||5.3±2.0 (2–9)||4.1±1.8 (1–7)||0.0240|
|CESD, mean±SD (range)||11.8±8.0 (0–31)||10.8±9.3 (0–28)||0.6743|
Fig. 3 represents the sensitivity and specificity of the measures of cognitive fatigue at the given cut off score. The (3*Stroop 60)/Stroop 180>1.075065 and (3*3DSAT 60)/3DSAT 180>1.150125 had a specificity of 100% in diagnosing cognitive fatigue with a sensitivity of 30.4% and 56.5% respectively. These tests were the most specific among the measures we used in this study. The (3*SDMCAT 60)/SDMCAT 180>1.10824 measure was the most sensitive with a sensitivity of 69.6%. The ROC curves of these individual measures are represented in Fig 4 and Fig 5. The area under curve for each of these measures is illustrated in Table 6. There was a significant positive correlation between both the electrophysiological measures and the clinical measures of cognitive fatigue used in this study (Table 7).
Sensitivity and specificity of various measures of cognitive fatigue.
ROC curves for electrophysiological measures of cognitive fatigue.
ROC curves for clinical measures of cognitive fatigue.
Area under curve for individual measures of cognitive fatigue.
|Measure||Area under curve||Standard error||95% CI|
|(3*Stroop 60)/Stroop 180||0.880||0.047||0.788–0.972|
|(3*mSDMT 60)/mSDMT 180||0.830||0.056||0.720–0.941|
|(3*SAT 60)/SAT 180||0.878||0.050||0.780–0.977|
|(3*SDMCAT 60)/SDMCAT 180||0.790||0.071||0.650–0.929|
|(3*3DSAT 60)/3DSAT 180||0.814||0.065||0.687–0.941|
Pearson correlation coefficients of the electrophysiological measures with the clinical measures of cognitive fatigue.
|(3*Stroop 60)/Stroop 180||0.6402||<0.0001||0.5773||<0.0001|
|(3*mSDMT 60)/SDMT 180||0.4046||0.0036||0.4044||0.0036|
|(3*SAT 60)/ SAT 180||0.5287||<0.0001||0.593||<0.0001|
|(3*SDMCAT 60)/SDMCAT 180||0.371||0.008||0.3591||0.0104|
|(3*3DSAT 60)/3DSAT 180||0.5055||0.0001||0.5179||0.0001|
In this study, we have derived evidence for all the three competing theories of cognitive fatigue. This could very well mean that cognitive fatigue in MS may be multifaceted. Our data supports the concept that as the time spent on a task increases, so does cognitive fatigue (the temporal fatigue hypothesis) – exemplified by the novel clinical measures used in this study - (3*Stroop 60)/Stroop 180, (3*SDMCAT 60)/SDMCAT 180, and (3*3DSAT 60)/3DSAT 180 – all of which indicated that if the time spent on a task is increased, the efficiency becomes lesser, in MS group compared to healthy controls. This finding corresponded to the results of the study by Sandry et al. Sandry et al., (2014) which also supported the temporal fatigue hypothesis. Although the methods used in that study were different and subjective measures of fatigue were also included, the results of their study are more or less similar to ours. In contrast to the MS patients, the performance efficiency increases for the healthy controls (indicated by scores <1 in these tests for cognitive fatigue) which may be due to the effects of learning. This also implies that impaired learning, in addition to cognitive fatigue may impair cognitive performance in MS patients.
To test the cognitive load hypothesis, we used two novel measures, (3*SDMCAT 60)/SDMCAT 180, and (3*3DSAT 60)/3DSAT 180 which presented a heavier cognitive load and hence had a higher mean scores (implying more cognitive fatigue) in the MS patient group (Table 4), compared to their simpler counterparts – the (3*mSDMT 60)/mSDMT 180 and (3*SAT 60)/SAT 180 respectively. Interestingly in the same table we can also note that the reverse is true for healthy controls – the scores on the ‘loaded’ versions of the tests were lesser than their simpler counterparts which could mean that normal healthy controls become more efficient in doing a cognitively more demanding task as time increases which might be the result of learning. This learning could be hampered by cognitive fatigue in patients with MS. Hence MS patients may not be able to keep up with their healthy counterparts in their performance as the cognitive demand of a task increases.
Though the (3*Stroop 60)/Stroop 180 were significantly different between the MS and healthy control group, the (3*mSDMT 60)/mSDMT 180 and (3*SAT 60)/SAT 180 were not which could imply that cognitive fatigue in MS could in fact be domain specific – giving some evidence in favour of the cognitive domain hypothesis. Larger studies are required to probe into this possibility.
The novel electrophysiological markers of cognitive fatigue used in this study - the P300L150: P300L50 and P300A50: P300A150 scores do correlate with the other clinical measures and hence may be used as a relatively quick test to identify MS patients who may be suffering from cognitive fatigue.
We had not included subjective measures of cognitive fatigue in this study since results from previous studies had proven that patients who tend to subjectively report more cognitive fatigue, do not show any corresponding decline in their performance assessed with standard tools. In many of our patients, we found the reverse also to be equally true – patients who self-reported little or no fatigue had a definite impairment on most of the scores. This discrepancy may be stemming from the inability of these individuals to accurately assess their own cognitive performance and also a lack of insight. The first step in any successful rehabilitatory measure is to make the individuals understand their own limitations. Therefore, it is of a greater significance to detect cognitive fatigue in these individuals and institute appropriate rehabilitatory or compensatory strategies (like breaking up a complex task into temporally separated smaller, less demanding tasks) which will improve their economic and social functioning and hence their quality of life. It is prudent to avoid maladaptive behaviours like straining more to complete a mentally demanding task in these individuals as this could be futile and frustrating both to these individuals and to their colleagues or caregivers owing to their fatigue related temporally declining cognitive performance.
Cognitive fatigue is a fairly common problem in MS patients – 46% of them had significant cognitive fatigue. All the novel clinical and electrophysiological measures used in this study are of value in detecting cognitive fatigue in MS patients. Data from this study also suggests that the sensitivity of the electrophysiological measures used in this study is relatively high, indicating that these may be used as a quick and efficient screening tool in detecting cognitive fatigue in MS patients.
This study supports all the three leading theories of cognitive fatigue in MS – the temporal fatigue, cognitive domain and the cognitive load hypotheses – indicating that cognitive fatigue may be a complex and multifaceted manifestation of diffuse white matter damage in MS.
- Ackerman and Kanfer, 2009 P.L. Ackerman, R. Kanfer. Test length and cognitive fatigue: an empirical examination of effects on performance and test-taker reactions. J. Exp. Psychol.. 2009;15(2):163 Crossref
- Bailey et al., 2007 A. Bailey, S. Channon, J. Beaumont. The relationship between subjective fatigue and cognitive fatigue in advanced multiple sclerosis. Mult. Scler.. 2007;13(1):73-80 Crossref
- Berrigan et al., 2013 L.I. Berrigan, J.-A. LeFevre, L.M. Rees, et al. Cognition in early relapsing-remitting multiple sclerosis: consequences may be relative to working memory. J. Int Neuropsychol. Soc.. 2013;19(08):938-949 Crossref
- Chaudhuri and Behan, 2000 A. Chaudhuri, P.O. Behan. Fatigue and basal ganglia. J. Neurol. Sci.. 2000;179(1):34-42 Crossref
- DeLuca et al., 2004 J. DeLuca, G.J. Chelune, D.S. Tulsky, et al. Is speed of processing or working memory the primary information processing deficit in multiple sclerosis?. J. Clin. Exp. Neuropsychol.. 2004;26(4):550-562 Crossref
- Fisk et al., 1994 J.D. Fisk, A. Pontefract, P.G. Ritvo, et al. The impact of fatigue on patients with multiple sclerosis. Can. J. Neurol. Sci.. 1994;21(1):9-14
- Freal et al., 1984 J. Freal, G. Kraft, J. Coryell. Symptomatic fatigue in multiple sclerosis. Arch. Phys. Med. Rehabil.. 1984;65(3):135-138
- Genova et al., 2009 H.M. Genova, F.G. Hillary, G. Wylie, et al. Examination of processing speed deficits in multiple sclerosis using functional magnetic resonance imaging. J. Int Neuropsychol. Soc.. 2009;15(03):383-393 Crossref
- Jensen et al., 2013 J.L. Jensen, D.A. Berry, T.A. Kummer. Investigating the effects of exam length on performance and cognitive fatigue. PLoS One. 2013;8(8):e70270 Crossref
- Kluger et al., 2013 B.M. Kluger, L.B. Krupp, R.M. Enoka. Fatigue and fatigability in neurologic illnesses proposal for a unified taxonomy. Neurology. 2013;80(4):409-416 Crossref
- Krupp and Elkins, 2000 L.B. Krupp, L.E. Elkins. Fatigue and declines in cognitive functioning in multiple sclerosis. Neurology. 2000;55(7):934-939 Crossref
- Krupp et al., 1988 L.B. Krupp, L.A. Alvarez, N.G. LaRocca, et al. Fatigue in multiple sclerosis. Arch. Neurol.. 1988;45(4):435
- Kurtzke, 1983 J.F. Kurtzke. Rating neurologic impairment in multiple sclerosis: an Expanded Disability Status Scale (EDSS). Neurology. 1983;33:1444-1452
- Minden et al., 2006 S. Minden, D. Frankel, L. Hadden, et al. The Sonya Slifka longitudinal multiple sclerosis study: methods and sample characteristics. Mult. Scler.. 2006;12(1):24-38 Crossref
- Parmenter et al., 2003 B.A. Parmenter, D.R. Denney, S.G. Lynch. The cognitive performance of patients with multiple sclerosis during periods of high and low fatigue. Mult. Scler.. 2003;9(2):111-118 Crossref
- Paul et al., 1998 R.H. Paul, W.W. Beatty, R. Schneider, et al. Cognitive and physical fatigue in multiple sclerosis: relations between self-report and objective performance. Appl Neuropsychol.. 1998;5(3):143-148 Crossref
- Polman et al., 2011 C.H. Polman, S.C. Reingold, B. Banwell, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann. Neurol.. 2011;69(2):292-302 Crossref
- Radloff, 1977 L.S. Radloff. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol. Meas.. 1977;1:1385-1401
- Sandry et al., 2014 J. Sandry, H.M. Genova, E. Dobryakova, et al. Subjective cognitive fatigue in multiple sclerosis depends on task length. Front. Neurol.. 2014;27(5):214
- Schwartz et al., 1996 C.E. Schwartz, L. Coulthard-Morris, Q. Zeng. Psychosocial correlates of fatigue in multiple sclerosis. Arch. Phys. Med. Rehabil.. 1996;77(2):165-170 Crossref
© 2016 Elsevier B.V., All rights reserved.