Multiple sclerosis (MS) is an inflammatory, neurodegenerative disease characterized by demyelination and axonal damage in the central and peripheral nervous systems. There are four courses of the disease and the most common one is a relapsing-remitting form (RRMS), in which the relapses of symptoms are followed by periods of remission with a different duration and a break between them. The courses differ in the type of attack by the immune system, clinical picture as well as neurological symptoms (Filippi et al., 2018; Lassmann, 2018). The MR image, which is partly used for diagnosis, shows focal lesions that are triggered by chronic inflammatory processes and cause symptoms of the disease, including mobility problems, pain, and cognitive decline (Filippi et al., 2018). On a neuronal level, diseased and structurally injured brain causes functional disruption of various neural networks. Therefore, functional networks (FNs) can be characterized with different activity in early and late stages of the disease. Deficits and disruption in one network are reported to cause hyperconnectivity in another network, which compensate for the loss in the disrupted network. It is believed hyperconnectivity represents an observable brain response to neural network disruption (Hillary and Grafman, 2017). Functional compensatory mechanisms are especially reported to be very common at the early, mild stages of the MS, which can limit first clinical manifestations of the disease until a certain threshold of damage is exceeded (Pantano et al., 2002; Audoin et al., 2005; Mainero et al., 2006; Basile et al., 2014; Droby et al., 2016; Castellazzi et al., 2018; Koubiyr et al., 2021). One of the most common (occurring in up to 80% of patients) symptoms in the early stages of MS is fatigue, influencing normal functioning (Simionia et al., 2007; Forwell et al., 2008; Gupta et al., 2013; Loy et al., 2017). Rudroff et al. (2016) propose the model, where fatigue in MS is described as “the decrease in physical and/or mental performance that results from changes in central, psychological and/or peripheral factors.” Nevertheless, it differs significantly from tiredness seen in healthy individuals or other chronic disease patients and many MS patients describe fatigue as the most debilitating disease sign (Bakshi, 2003; Kos et al., 2007; Induruwa et al., 2012). The phenomenon of fatigue in MS can be understood on the level of state fatigue, which relates to psychological state in the specific moment of time or fatigue as a trait. Enoka et al. (2021) divide trait fatigue into perceived fatigability and objective fatigability. The former can be operationalized from capacity to perform past, present, and future activities, while the latter refers to limits of current actions, manifested in voluntary activations and contractile function.

The mechanisms of fatigue could be divided into primary ones such as involvement of immune system and altered patterns of brain activation through lesions or secondary mechanisms, which results from related conditions like sleep disorders or depression (Braley and Chervin, 2010). Nevertheless, the origin and mechanism of this symptom are still poorly understood. However, fatigue in MS is thought to be mostly linked to age, disability, and disease duration (Colosimo et al., 1995; Ghajarzadeh et al., 2013; Loy et al., 2017), with some suggestion that disability is a driving factor (Kroencke et al., 2000).

The researchers have been already looking for structural and functional correlates of fatigue in MS, however, metaanalyzes show that most of the studies focus on MS patients with moderate or severe disability status (Loy et al., 2017; Moss-Morris et al., 2021). So far, Wilting et al. (2015) using voxel-based morphometry revealed thalamic alterations associated with fatigue in the early stage of relapsing-remitting MS. Functional abnormalities of the brain could be distinguished by resting-state fMRI (rs-fMRI), which enables studying functional interactions in the state of rest, without any explicit task performance (Smith et al., 2013). These interactions and their alterations are determined in healthy individuals in various conditions and clinical approaches, including patients with MS. The previous research on MS indicated diminished synchronization in the cerebellum, thalamus, or insula with other brain regions (Zhou et al., 2013; Tona et al., 2014). Moreover, the alterations in functional connectivity (FC) in MS are correlated with the disability progression (Faivre et al., 2016; Tommasin et al., 2020). Zhou et al. (2016) also revealed disturbances in brain atrophy, a new approach to investigate brain temporal dynamics, in patients with MS compared to healthy subjects. Several studies associate resting-state FC with MS-related fatigue (e.g., Bisecco et al., 2017; Jaeger et al., 2018; Stefancin et al., 2019), demonstrating divergent results.

The disturbances of FC were so far revealed mainly in DMN (Bonavita et al., 2011, 2016; Zhou et al., 2014) and salience network (SN) (Cruz-Gómez et al., 2013; Manca et al., 2019). Moreover, the previous research revealed altered amplitude of low-frequency fluctuation (ALFF) in, i.e., motor, cerebellar, or frontal regions (Liu et al., 2011; Mansoory et al., 2018; Plata-Bello et al., 2018).

The aim of the current study is to determine the fatigue-related FC disturbances in default mode network (DMN), SN, and sensorimotor network (SMN) as well as whole brain ALFF and fractional amplitude of low-frequency fluctuations (fALFF) in the mild stage of RRMS. Based on the previous literature, we hypothesize hyperconnectivity in DMN, SN as well as SMN as a sign of compensation mechanisms in networks which are thought to be disrupted during the progression of MS. Moreover, we hypothesize altered ALFF, fALFF and graph measures in frontal, motor, and cerebellar regions.

Cluster 1 consisted of 53 patients (20 men) aged from 20 to 56 (M = 35.06; SD = 7.87). Mean FSS score (M = 2.41; SD = 0.74), EDSS score (M = 1.15; SD = 0.79), years since onset of first MS symptoms (M = 8.06; SD = 4.73) as well as years since the start of MS related treatment (M = 5.43; SD = 3.23) were calculated. Cluster 2 consisted of 48 patients (17 men) aged from 25 to 59 (M = 38.46; SD = 7.91). Mean FSS score (M = 4.64; SD = 0.77), EDSS score (M = 1.54; SD = 1.06), years since onset of first MS symptoms (M = 9.81; SD = 4.62), and years since the start of MS related treatment were calculated (M = 6.98; SD = 3.5). Clusters differed significantly in age [t_(2,99) = 2.16; p = 0.033] and the older patients were placed in the Cluster 2. Moreover, patients from Cluster 2 also obtained significantly higher scores in FSS [t_(2,99) = −14.79; p < 0.001], EDSS [t_(2,99) = 2.083; p = 0.04], and were treated for MS longer than participants from Cluster 1 [t_(2,99) = 2.308; p = 0.023]. Clusters showed no significant difference in gender χ² = (1, n = 101) = 0.058; p = 0.809 and years from the first onset of MS symptoms t_(2,99) = 1.882; p = 0.063. All demographic values are showed in Table 1.

A one-way ANCOVA was conducted to investigate potential between-group differences in FC of DMN, SN, and SMN associated with the level of fatigue. The analysis revealed statistically significant difference between patients in Cluster 1 and Cluster 2 in posterior SN, while controlling for age, disability level, and duration of pharmacological treatment [F_(2,90) = 6.093; p = 0.016]. Post hoc comparisons using Bonferroni test indicated that the mean FC values for the Cluster 1 (M = 0.401; SD = 0.115) was significantly different from mean FC values for Cluster 2 (M = 0.457; SD = 0.113). These results suggests that the FC between brain areas among posterior SN was higher in Cluster 2 in which the fatigue level was increased as compared to Cluster 1 and that this result is independent of age, disability level, and duration of pharmacological treatment. A one-way ANCOVA showed no significant differences between Cluster 1 and Cluster 2 in FC among anterior SN [F_(2,91) = 1.858; p = 0.176], dorsal [F_(2,91) = 1.353; p = 0.248], and ventral [F_(2,91) = 0.906; p = 0.344] parts of the DMN as well as SMN [F_(2,91) 0.590; p = 0.444]. Furthermore, two sample t-test showed statistically significant between-group differences in fALFF [t_(2,99) = 3.17; p = 0.027]. The threshold was set at the p-value 0.001 with cluster size of 28 voxels. Patients with higher fatigue level had lower fALFF in right middle frontal gyrus as well as right superior frontal gyrus. No differences were found in ALFF as well as graph measures (p > 0.05). All of the fMRI results are FDR corrected with p < 0.05. Visualization of above results can be found in Figures 1, 2.

The aim of the current study was to investigate the association of FN architecture in DMN, SN, and SMN as well as whole-brain ALFF and fALFF with the level of fatigue in the mild stage of RRMS.

The investigation revealed higher average FC in posterior SN in Cluster 2 in comparison with Cluster 1. The SN play an enormous role in attentional processes such as its distribution and capacity (Manca et al., 2019), enabling productive information processing as well as faster responses to stimuli. At the same time, people with chronic fatigue syndrome are coherently described as the ones with significant deficits in attention (Ray et al., 1993; Holgate et al., 2011). Accordingly, subjects from Cluster 2, who declare to be almost twice as much tired as participants from Cluster 1, manifest decreased attention, compared to the group with the lower scores of fatigue. Therefore, higher FC in posterior SN in RRMS patients might be a compensatory mechanism which enables the patients to cope with deficits of attention. Many studies point to the compensation mechanisms in MS patients, especially in the early stages (Pantano et al., 2002; Audoin et al., 2005; Mainero et al., 2006; Basile et al., 2014; Droby et al., 2016; Castellazzi et al., 2018; Koubiyr et al., 2021). Hyperconnectivity is thought to represent an observable brain response to neural network disruption caused by diseased and structurally injured brain (Hillary and Grafman, 2017). However, only short-term hyperconnectivity is considered to be beneficial, while long-term hyperconnectivity can have bad consequences for human brain. Increased brain metabolism, caused by a local hyperconnectivity, is a predictor of higher amyloid beta (Aβ) deposition, which is a well-known signature of Alzheimer’s disease. In conclusion, chronic hyperconnectivity, performing a function of compensatory mechanisms, can lead to a serious progressive neurodegeneration (Vlassenko et al., 2010; Myers et al., 2014; Hillary and Grafman, 2017). Noteworthy, the main region of the posterior SN is posterior insula, which is widely known to be involved in interoceptive awareness, understood as a capability to knowingly perceive inner bodily signals (Kuehn et al., 2016). According to Craig et al. (2000), changes in the activity of posterior insula correlate with the graded intensity of interoceptive stimuli, while stimulation of the same region of interest results in interoceptive and somatic sensations (Penfield and Faulk, 1955). Moreover, posterior insula is thought to play a crucial role in nociception (Frot et al., 2014; Segerdahl et al., 2015). Higher nociception in MS patients stays in line with the literature, as chronic neuropathic pain is frequent and very characteristic symptom of this neurodegenerative disease (Yilmazer et al., 2020). Current study is also congruent with the research of Chen et al. (2020), who observed increased FC in the parts of insula which were described as an interoceptive hub. What is more, the study concludes that hyperconnectivity of interoceptive network is associated with experiencing of the cognitive fatigue.

Another result from the current study revealed smaller fALFF in RRMS patients with higher fatigue score in the cluster located in right middle frontal gyrus and right superior frontal gyrus. In the study of Sepulcre et al. (2009) lesion load in the exact same regions was positively correlated with the fatigue score. Authors concluded that lesions in aforementioned regions are associated with the disruption of FNs responsible for cognitive and attention processes. Other studies also point out the relation of middle frontal gyrus and fatigue in MS (Filippi et al., 2002; Calabrese et al., 2010). Noteworthy, both right middle frontal gyrus as well as the right superior frontal gyrus are thought to be located at the borderline between DMN and cognitive executive network (Young et al., 2017), while the SN is reported to modulate the switch between aforementioned two networks (Menon, 2015). Above results are very promising, considering hyperconnectivity in SN in MS patients with higher fatigue.

Noteworthy, brain regions with significantly altered FC as well as fALFF are anatomically or functionally connected to striatal-thalamic-frontal network, commonly known as the fatigue network (Genova et al., 2013). Genova et al. (2013) with the use of cognitively fatiguing task and FSS revealed that striatum as well as thalamus, frontal regions and its interconnected regions play significant role in perceived fatigability in MS patients. Interestingly, the SN, which had higher FC in patients with higher fatigue, consists regions which shows intrinsic connection to striatum and thalamus (Menon, 2015). The same group of patients manifested smaller fALFF in middle and superior frontal gyrus, which overlap with the striatal-thalamic-frontal network. Therefore, our results are partly congruent with the results of Genova et al. (2013) pointing on neural correlates of perceived fatigability in MS.

The current study revealed that MS patients with higher fatigue level were characterized with significantly higher disability status, age as well as longer pharmacological treatment. Above results are congruent with various previous studies (Colosimo et al., 1995; Kroencke et al., 2000; Ghajarzadeh et al., 2013) as well as the meta-analysis of Loy et al. (2017), who reports that MS patients with higher EDSS and longer disease duration suffer from fatigue more than patients with lower EDSS. Aforementioned questionnaire results, which stays in line with other studies in the field, strengthen the fMRI results on the neural basis of fatigue in mild stage MS.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Institute of Applied Psychology Ethics Committee of the Jagiellonian University. The patients/participants provided their written informed consent to participate in this study.

MF, TM, MW, AMS, BB, AC, and AZ: conception and design of the work. AMS, BB, AC, and AZ: data acquisition. AMS, BB, and AC: rs-fMRI data analysis. AMS, BB, MF, and TM: interpretation of the results. AMS, BB, AC, AZ, MW, AS, MM, KN, and KZ-W: drafting the work. AMS, BB, AC, AZ, MF, TM, MW, AS, MM, KN, and KZ-W: revising. MW, AS, MM, KN, and KZ-W: patients recruitment and diagnosis, revising the manuscript critically and final approval of the version to be published. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.