The Relationship Between Fatigue And A Clinically Accessible Measure Of Switching in Individuals With Multiple Sclerosis

Abstract

Objective: 

We examined whether fatigue in multiple sclerosis (MS) is linked to switching processes when switching is measured by the Trail Making Test (TMT).

Multiple sclerosis is a neurological disease that affects a person's movement, vision, language, and cognitive abilities. However, even with this disease, we can still actively cope and improve our memory.

First, we can improve our memory through exercise. Research shows that exercise is critical to the health and survival of neurons. For people with MS, moderate exercise is necessary to keep their body and brain healthy. For example, regular aerobic exercise, such as walking, jogging, or swimming, can enhance blood circulation and neuron connections in the brain.

Secondly, we can try to master new skills, which is a very important challenge for people's brains. Learning new skills stimulates neuronal connections and helps our brains store information more efficiently. For example, trying to learn a new language, learning a musical instrument, or trying to learn a craft can exercise our brains and improve our memory.

In addition, we can also change our eating habits. Certain nutrients found in our meals can help our brains function more healthily. For example, Omega-3 fatty acids enhance the functioning of the nervous system and help our brains process information more efficiently.

In summary, although having MS can affect our memory and cognitive abilities, we can still improve our memory and keep our brains healthy through appropriate exercise, learning new skills, and changing our eating habits. It can be seen that we need to improve our memory. Cistanche deserticola can significantly improve memory because Cistanche deserticola is a traditional Chinese medicinal material with many unique effects, one of which is to improve memory. The efficacy of minced meat comes from the various active ingredients it contains, including acid, polysaccharides, flavonoids, etc. These ingredients can promote brain health in various ways.

Click Know Short-term Memory how to improve

Method:

Eighty-three participants with MS were administered a battery of standardized tests of switching, working memory, and processing speed. Ordinary least squares regression models were used to estimate the association between fatigue severity and switching above and beyond attention, working memory, and processing speed.

Results: 

We found a negative association between TMT performance and fatigue severity score. When measures of processing speed and working memory were included in the model, the switching measure continued to uniquely contribute to fatigue severity.

Conclusions: 

There may be a unique relationship between fatigue and switching processes identifiable by clinical measures of switching. Future research should continue to investigate this relationship by using both behavioral and neural markers to test models of fatigue to eventually identify specific intervention targets.

Keywords: 

Multiple sclerosis; Assessment; Executive functions

Introduction

Fatigue is observed in 27%–95% of individuals with neurological syndromes including multiple sclerosis (MS), Parkinson’s disease, stroke, myasthenia gravis, and traumatic brain injury (Kluger, Krupp, & Enoka, 2013). In MS, fatigue is a common and debilitating symptom, negatively impacting the performance of daily activities, work, and quality of life (Krupp, Serafin, & Christodoulou, 2010). Despite its prevalence and negative impact, the precise mechanisms that contribute to fatigue remain unclear. As such, the effective mechanisms and limitations of typically recommended physical exercise or pharmaceuticals for fatigue management also remain unclear (Heine, van de Port, Rietberg, van Wegen, & Kwakkel, 2015; Nourbakhsh et al., 2021). To better assess and treat fatigue, we must clarify the underlying mechanisms of the symptom.

In the cognitive neuroscience literature, researchers suggest that dysfunction within the frontostriatal circuit is central to fatigue (Chaudhuri & Behan, 2000; Dobryakova, DeLuca, Genova, & Wylie, 2013). Prefrontal systems involved in the frontostriatal circuit (e.g., prefrontal cortex; PFC) are also highlighted in the costs and benefits framework described by Boksem and Tops (2008), which posits that fatigue is the result of a mismatch between the amount of effort and reward required and received for a task. Support for these views comes from research that demonstrates a relationship between fatigue on cognitive tasks and cognitive control processes (Shenhav, Cohen, & Botvinick, 2016). 

Cognitive control functions can be broadly divided into different components, including switching (Szczepanski & Knight, 2014). Some findings in fatigue research highlight that only attention and vigilance tasks are most related to fatigue (Hanken, Eling, & Hildebrandt, 2015), whereas others have shown that tasks involving cognitive control functions resulted in increased response times or more errors with increasing fatigue (Lorist et al., 2000). Importantly, fatigue research showed that fatigued individuals demonstrated more performance deficits on executive functioning tasks (e.g., Wisconsin Card Sorting Test) compared with tasks that required individuals to maintain and reproduce information (e.g., digit span) (van der Linden, Frese, & Meijman, 2003).

Prefrontal regions in the brain associated with cognitive control are considered essential to task-switching ability (Herd et al., 2014). Switching tasks relies on maintaining task-rule representations to correctly respond to a stimulus as well as inhibitory control to prevent executing the wrong task rule (Monsell, 2003). Along with behavioral research showing associations between fatigue and cognitive control processes, neuroimaging research also shows an association between fatigue and increased cerebral activation in frontostriatal regions (Dobryakova et al., 2013). Given the associations between fatigue and cognitive control processes in both behavioral and neuroimaging studies, we were interested in the specific link between switching and fatigue, using a common and clinically accessible neuropsychological measure of switching.

In neuropsychological assessment, the Trail Making Test (TMT) is one of the most widely used instruments measuring cognitive processing speed and cognitive control (Strauss et al., 2006). TMT A requires mainly visuoperceptual abilities, attention, and graphomotor speed, whereas TMT B primarily reflects working memory and task-switching ability. The difference between these subtests (B-A) has been shown to minimize visuoperceptual and working memory demands, providing a relatively pure indicator of cognitive control abilities, specifically switching (Sánchez-Cubillo et al., 2009). The TMT is also strongly associated with prefrontal regions as well. Specifically, research suggests that the left PFC is a critical region for the processing demands of TMT (Miskin et al., 2016) and is strongly associated with cognitive control (Szczepanski & Knight, 2014). 

Moreover, the TMT is simple to administer and typically takes <5 min to complete while maintaining sensitivity comparable to other longer and more complex neuropsychological measures of cognitive control functioning (e.g., Wisconsin Card Sorting Test). Together, the TMT’s psychometric reliability and sensitivity to cognitive control dysfunction in individuals, and its administrative simplicity, make it a clinically efficient behavioral measure of switching. A specific relationship between switching processes and fatigue could suggest that prefrontal regions mediating cognitive control and switching processes may be specifically linked to fatigue. Moreover, the relationship between switching and fatigue could be captured using a readily available and utilized neuropsychological measure of switching.

The current study examines the relationship between a clinical measure of switching and fatigue. We hypothesized that performance on a switching measure would be associated with increased fatigue. Given the role of processing speed and working memory in cognitive control and since both are implicated in MS, we further tested whether switching is a specific cognitive control process linked to fatigue. We predicted that a clinical measure of switching would be associated with fatigue, above and beyond clinical measures of working memory and processing speed.

Methods

Participants and procedures

Participants were selected across existing studies of neuropsychological functioning in MS. MS participants were recruited at Drexel University from the greater Philadelphia area and through an outpatient clinic of the Comprehensive Multiple Sclerosis Center at Jefferson University Hospital. The combined dataset included 84 individuals with clinically definite MS verified by a review of medical records submitted by a treating physician. Notably, at the time of recruitment, participants met the 2011 McDonald criteria for the diagnosis of MS. Participants ranged from 18–60 years of age, had been diagnosed with MS (all subtypes) for at least 1 year, were at least 30 days from their most recent exacerbation, and reported no current corticosteroid or stimulant use. All participants had no other neurological disease, psychiatric disease, psychosis, substance abuse, or current use of medications shown to adversely affect cognition at the time of enrollment.

This study’s protocol received prior approval by the institutional review board responsible for ethical standards at Drexel University. All participants gave written informed consent before participating in this study. The research visits across studies lasted 2 h and included demographic and neuropsychological evaluation.

Fatigue assessment

We used the Fatigue Severity Scale (FSS) to measure fatigue in participants. The FSS is a 9-item scale validated in various clinical, behavioral, and neuroimaging studies to investigate fatigue and, specifically, fatigue in MS (Krupp, 1989). Each of the nine items is completed on a rating scale of one to seven, and a final score is the total of the nine ratings. Subjects were asked to rate the fatigue they had experienced over the previous week.

Neuropsychological assessment

The following specific neuropsychological measures were selected among those commonly available across prior studies to test our specific hypotheses. The measures of specific interest included switching, attention, working memory, and processing speed. Parts A and B of the TMT were administered and recorded in total time in seconds. The difference score (B-A) measures switching ability (Sánchez-Cubillo et al., 2009) and served as the primary independent variable of interest. The Paced Auditory Serial Addition Test (PASAT) 2-s measures sustained attention and information processing speed. The PASAT is also conventionally thought to have working memory demands. The Symbol Digit Modalities Test (SDMT) (oral version) measures information processing speed and working memory and offers another modality to control for working memory and processing speed. To mitigate the potential for practice effects, participants had not participated in any neuropsychological testing that included any of the measures of interest in at least 1 year.

Statistical analyses

All analyses were conducted in R version 3.6.2 (R: The R Project for Statistical Computing, 2020). All data were screened for normality and determined to be within acceptable limits. To determine which covariates should be included in our analyses and given the overlap in clinical measures on cognitive domains, Spearman’s correlations were used to investigate the relationships between age, FSS score, TMT difference, PASAT, and SDMT measures. We constructed two linear regression models to evaluate the magnitude of the relationship between TMT difference and fatigue severity, with FSS as the dependent variable. The first regression included only TMT difference as the predictor variable. The second regression was created with PASAT and SDMT as additional predictors to ensure the effect of TMT difference on fatigue severity is not due to any measured aspect of processing speed or working memory aspects of TMT.

Results

Of the original 84 participants, data from one participant were not included in the main analyses due to performance on a single task (8 SD above the mean) violating assumptions to the statistical test. Therefore, data from 83 individuals (63 female) were analyzed. Patient demographics and disease characteristics are described in Table 1.

The mean FSS score was 40.56 (±13.64, range 10–63). We calculated Spearman’s correlations to examine the relationship between fatigue severity, age, and cognitive measures. FSS score was not significantly correlated with TMT A performance, but FSS was significantly correlated with TMT B performance (0.22, p = .03). FSS was significantly correlated with TMT difference score (0.23, p < .01). TMT difference score was significantly correlated with TMT A (0.12, p < .001), TMT B (0.85, p < .001), PASAT (−0.36, p < .01), and SDMT (−0.43, p < .001) performance. TMT A was significantly correlated to TMT B (0.58, p < .001), PASAT (−0.39, p < .001), and SDMT (−0.64, p < .001) performance. TMT B was also significantly correlated with PASAT (−0.51, p < .001) and SDMT performance (−0.66, p < .001). PASAT performance and SDMT performance were also significantly correlated, (0.51, p < .001). We did not find any other significant correlations between measures. A summary of neuropsychological measures is reported in Table 1.

Regression results investigating the relationship between FSS and TMT differences are reported in Table 2. Consistent with our hypothesis, TMT difference was significantly associated with FSS score, such that an increase in time in TMT-B versus TMT-A (measured in seconds) was associated with a 0.21 (95% confidence interval [CI]: 0.05–0.36) point increase in FSS score, p = .009, R2 = 0.08, F(1, 81) = 7.2, p = .008. Given the correlations between TMT difference, PASAT, SDMT, and FSS, we performed a second regression with PASAT and SDMT as additional predictors. In this model, PASAT and SDMT were not significant; however, the TMT difference score maintained statistical significance (β = 0.2, p = .03, 95% CI: 0.02–0.37) (see Table 2).

Discussion

The current study examined the relationship between a clinical measure of switching and fatigue in a sample of individuals with MS. Consistent with our hypothesis, our findings showed that switching, as measured by the TMT, was associated with increased fatigue severity. When we included measures of processing speed and working memory in our analyses, we found that switching was uniquely associated with fatigue.

Overall, the current findings demonstrate a link between switching and fatigue. Switching behavior has sustained cognitive control demands that may be uniquely related to fatigue (Lorist et al., 2000). Many studies show that the clinical population exhibits increased activation in frontal and striatal regions associated with cognitive control during demanding cognitive tasks. Therefore, if the white matter disease process of MS affects cognitive control processes, which are often recruited during cognitive tasks and associated with a sense of felt effort, and cognitive control is persistently recruited to a greater degree to support performance, fatigue could result. The dysfunction would then explain the increased fatigue experienced by individuals with MS and potentially other neurological disorders during effortful tasks that require cognitive control. Notably, the possibility that fatigue is fundamentally related to neural dysfunction in circuits that regulate effortful cognitive control, specifically during switching tasks, requires additional research. There may be other circuits or processes more specifically linked to fatigue. In their review, Hanken and his colleagues (2015) suggest fatigue impacts performance on alertness and vigilance tasks more than memory, processing speed, language, or visuospatial tasks. Notably, their review did not explicitly examine cognitive control as a domain of interest. Although attention and vigilance are constructs necessary for successfully executing cognitive control processes, including switching, future studies should continue to test whether fatigue is uniquely associated with domains that do not require sustained cognitive control demands.

Our findings also support and extend prior work by showing a relationship between fatigue and an objective measure of switching specifically, using the TMT difference score. Clinicians and researchers often rely on subjective reports of fatigue; however, subjective reports can be limited in that they rely on self-evaluations and provide little information about the real or perceived causes of an individual’s fatigue. Therefore, a performance-based clinical tool could substantiate self-report measures of fatigue by offering an objective measure related to reports of fatigue. Notably, objective performance is not consistently correlated with self-reported measures in fatigue studies (Benoit et al., 2019). One explanation for the inconsistency between self-reported fatigue and task performance is the use of tasks that do not target the fatigue process. Although the current study revealed a small relationship between the TMT difference score and fatigue, our findings are consistent with Park and Larson’s (2015) study which, to our knowledge, is the only other study to examine the relationship between TMT difference score and fatigue. The relationship between performance on the TMT (Johansson, Berglund, & Rönnbäck, 2009; Park & Larson, 2015) or neuropsychological tests more broadly (Vercoulen et al., 1998) and fatigue is mixed in the literature. Our study looked at the specific relationship between cognitive switching and fatigue and findings supported those in the literature that showed an association between switching and fatigue. Moreover, we examined whether switching, as measured by the TMT, had a unique relationship with fatigue relative to other closely related cognitive processes. 

The relationship we found between TMT difference score and fatigue compared with performance on the PASAT and SDMT suggests that there is a unique relationship between processes involved in TMT and fatigue. Within the TMT subtests, the relationship we found between TMT-B and switching, and not TMT-A, could suggest that TMT-B may be driving the effect we found for the TMT difference score. In either case, these findings further support the relationship between switching and fatigue. Given a relationship between performance on a switching measure and reported fatigue, dysfunction in processes required for the successful execution of an objective measure (e.g., cognitive control for switching in TMT) may be central to the individual’s fatigue. If the opposite is true, the underlying mechanisms of fatigue for that individual may come from another source. Nevertheless, identifying an appropriate behavioral task that demonstrates objective performance consistent with subjective fatigue would have significant implications for assessment and intervention techniques.

There are several limitations to this study. Foremost, we were limited by the retrospective nature of the study and the data available to us. For instance, we did not have MS functional status or psychosocial measures (e.g., depression measures) available to include in our analyses. Although participants were excluded based on active depression, it is important for future research to include and account for the emotional state of participants and the impact it may have on performance and the interpretation of the results (Strober & Arnett, 2005). In addition, fatigue is a multifaceted construct, generally thought to be made up of physical and cognitive dimensions. Thus, cognitive neuroscience references to the frontostriatal circuit and fatigue are primarily within the context of cognitive fatigue. Questions on the FSS do not clearly distinguish cognitive versus physical fatigue and instead reflect a more global assessment of the construct. Although various fatigue measures are highly correlated with FSS, future studies may consider the use of a measurement tool specific to or separately measuring the different dimensions of fatigue. Another limitation is the reported effects are specific for patients with MS, therefore future studies should include a control group to help distinguish whether the relationship between switching and fatigue is specific to MS or for individuals irrespective of MS diagnosis.

The current study utilized a commonly administered neuropsychological measure of switching to examine its relationship with fatigue. Although our findings are largely correlational, results support and extend previous studies that use more elaborate switching tasks to study fatigue. Our results are also consistent with models that suggest that cognitive fatigue is central to prefrontal processes that mediate task-switching performance. The current results suggest that switching may be specifically implicated in the experience of fatigue, can be captured using a clinically efficient neuropsychological tool, and that an underlying deficit in switching processes may be central to fatigue experienced in MS. Future studies should explicitly test the neurobiology of other cognitive mediators of fatigue, including motivation and reward systems, as has been suggested by several researchers (Boksem & Tops, 2008; Hockey, 2011). Future studies should also expand on the relationship between switching, effort, reward, and fatigue to identify specific behavioral and neurobiological processes to assess and treat fatigue in the future.

Funding

This work was supported by the National Multiple Sclerosis Society under Grant [RG 5164A4/2] to MTS. FE acknowledges support from the National Institutes of Health grant [1F31NS122411-01]. JDM acknowledges support from National Institutes of Health grants [DP5-OD-021352-01, R01-DC-16800-01A1, R01-DC-014960-01A1, R01-AG-059763], and Department of the Army grant [PRMRP 12902164].

References

1.Benoit, C.-E., Solopchuk, O., Borragán, G., Carbonnelle, A., Van Durme, S., & Zénon, A. (2019). Cognitive task avoidance correlates with fatigue-induced performance decrement but not with subjective fatigue. Neuropsychologia, 123, 30–40.

2.Boksem, M. A. S., & Tops, M. (2008). Mental fatigue: costs and benefits. Brain Research Reviews, 59(1), 125–139.

3. Chaudhuri, A., & Behan, P. O. (2000). Fatigue and basal ganglia. Journal of the Neurological Sciences, 179(S 1–2), 34–42.

4.Dobryakova, E., DeLuca, J., Genova, H. M., & Wylie, G. R. (2013). Neural correlates of cognitive fatigue: cortico-striatal circuitry and effort–reward imbalance. Journal of the International Neuropsychological Society, 19(8), 849–853.

5.Hanken, K., Eling, P., & Hildebrandt, H. (2015). Is there a cognitive signature for MS-related fatigue? Multiple Sclerosis Journal, 21(4), 376–381.

6.Heine, M., van de Port, I., Rietberg, M. B., van Wegen, E. E., & Kwakkel, G. (2015). Exercise therapy for fatigue in multiple sclerosis. Cochrane Database of Systematic Reviews. (9), 1–126.

7. Herd, S. A., O’Reilly, R. C., Hazy, T. E., Chatham, C. H., Brant, A. M., & Friedman, N. P. (2014). A neural network model of individual differences in task switching abilities. Neuropsychologia, 0, 375–389.

8. Hockey, G. R. J. (2011). A motivational control theory of cognitive fatigue. In P. L. Ackerman (Ed.), Cognitive fatigue: Multidisciplinary perspectives on current research and future applications (pp. 167–187). American Psychological Association, Washington, DC.

9.Johansson, B., Berglund, P., & Rönnbäck, L. (2009). Mental fatigue and impaired information processing after mild and moderate traumatic brain injury. Brain Injury, 23(13–14), 1027–1040.

10.Kluger, B. M., Krupp, L. B., & Enoka, R. M. (2013). Fatigue and fatigability in neurologic illnesses: proposal for a unified taxonomy. Neurology, 80(4), 409–416.

For more information:1950477648nn@gmail.com