A randomized comparative feasibility study of neuromodulation and cognitive training for post-COVID fatigue

Abstract

Fatigue and cognitive deficits are common and disabling symptoms in patients experiencing post-COVID condition. This randomized parallel study aimed to evaluate the effects of transcranial direct current stimulation (tDCS) over the primary motor cortex combined with cognitive training (M1 + CT), compared to tDCS over the dorsolateral prefrontal cortex with cognitive training (DLPFC + CT), on fatigue, cognition, and other clinical symptoms in post-COVID. Sixty-three patients completed the treatment (n = 32 in the M1 + CT group and n = 31 in the DLPFC + CT group) with a mean age of 47 years and an average symptom duration of 32 months. Both groups underwent comprehensive neuropsychological and clinical evaluations, including ecological momentary assessments of fatigue, at baseline, post-treatment, and one-month follow-up. The Fatigue Severity Scale (FSS) was used as the primary endpoint. Patients were randomly assigned to the M1 + CT or DLPFC + CT groups and received 15 sessions of tDCS administered concurrently with adaptive CT. The M1 + CT group showed a slightly higher efficacy in reducing fatigue and improving sleep quality than the DLPFC + CT group. Both groups demonstrated significant improvements in cognition, anxiety, depression, pain, and sleep quality. These improvements were sustained over time. These findings indicate that tDCS combined with cognitive training is a feasible, safe, and effective approach for reducing fatigue and enhancing cognition in post-COVID patients. The results highlight the potential of brain stimulation and cognitive training to alleviate fatigue and cognitive impairment in post-COVID, warranting further confirmation through additional randomized controlled trials.

Trial registration: ClinicalTrials.gov NCT05753202.

Introduction

Long COVID, or post-COVID condition affects individuals with a history of SARS-CoV-2 infection, where symptoms persist for over 3 months¹. Approximately 5–10% of individuals who survive to COVID-19 develop post-COVID, with a portion not achieving full recovery^2,3. Among these, fatigue and cognitive dysfunction are the most frequent and disabling⁴. Fatigue, affecting 35–60% of patients⁵, manifests both physically and cognitively^6,7. Common cognitive deficits include impairments in attention, working memory, and processing speed. These symptoms have been linked to structural and functional brain changes^{7, 8–9}, negatively impacting patients’ quality of life and work capacity⁶. Given the burden of these symptoms, and the persistence of these symptoms over time^10,11 effective treatments are urgently needed.

Non-invasive brain stimulation techniques have shown promising results improving these symptoms in post-COVD patients¹². In particular, transcranial direct current stimulation (tDCS) is a non-invasive, portable, and safe neuromodulation technique¹³ that has demonstrated efficacy in improving fatigue in conditions such as multiple sclerosis¹⁴, fibromyalgia¹⁵, and other autoimmune disorders¹⁶. However, limited research exists on the effectiveness of tDCS in post-COVID. A previous randomized sham-controlled study from our group showed that stimulation over the left dorsolateral prefrontal cortex (DLPFC) compared to sham-group improved physical fatigue in post-COVID patients, but there were no significant effects on cognitive fatigue, cognition, or other symptoms¹⁷. Another study reported improvements in cognitive fatigue following stimulation over the left primary motor cortex (M1) compared to sham-group, though improvements in cognition were not observed, and no assessment of ecological fatigue was conducted¹⁸.

Cognitive training (CT) has shown efficacy in improving cognition and can complement tDCS¹⁹. CT has been effective in various disorders, such as mild cognitive impairment²⁰, autoimmune diseases²¹, and neurodegenerative conditions like Parkinson’s disease²². Combining tDCS with CT may enhance cognitive benefits, especially in domains frequently impaired in post-COVID^{23, 24–25}.

Building on these findings, more comprehensive treatment strategies targeting both fatigue and cognition are needed for post-COVID patients. Based on previous evidenced effects of tDCS compared to sham on improving fatigue in post-COVID^17,18, the present study aimed to design and evaluate the efficacy of a more comprehensive treatment aiming to improve not only fatigue but also cognition. In this context, this study aimed to investigate and compare the effects of tDCS over two different targets— M1 and DLPFC—combined with adaptive CT in post-COVID patients.

The primary objective was to compare the effects of tDCS over M1 combined with CT (M1 + CT) versus tDCS over the DLPFC combined with CT (DLPFC + CT) on physical fatigue in post-COVID. Secondary objectives included assessing the effects on cognitive fatigue, cognition, depression, anxiety, pain, sleep quality, and quality of life, as well as assessing the feasibility and safety of both interventions.

Results

Participants

Patients had a mean age of 47.83 years and 80.95% were female, with a mean of 16.20 years of education. These patients had a mean of 32 months of evolution since the infection. Both groups were similar in terms of age, sex and education and premorbid risk factors (Table 1).

Table 1. Sociodemographic and clinical characteristics.

	Total (n = 63)	M1 + CT (n = 31)	DLPFC + CT (n = 32)	T/χ2	p-value
Age (years)	47.83 (8.13)	47.35 (8.45)	48.28 (8.13)	0.445	0.658
Sex (Females) (n,%)	51 (80.95%)	26 (83.87%)	25 (78.12%)	0.337	0.561
Education (years)	16.20 (3.78)	15.90 (4.32)	16.28 (3.07)	0.402	0.690
Evolution (months)	32.02 (7.91)	31.10 (8.73)	34.56 (19.83)	0.906	0.368
Premorbid risk factors
Hypertension (n,%)	5 (7.9%)	2 (6.5%)	3 (9.37%)	0.184	0.668
Diabetes (n,%)	3 (4.8%)	1 (3.22%)	2 (6.25%)	0.318	0.573
Dyslipemia (n,%)	16 (25.0%)	7 (22.6%)	9 (28.12%)	0.255	0.613

Feasibility and safety

Among the patients who started the treatment, 94% completed the trial. Of those who declined to continue, two patients were reinfected with COVID-19, one had scheduling incompatibility with work and another did not tolerate the CT.

Regarding side effects, 14.28% of the patients reported adverse effects. Specifically, 7.9% reported headache and 6.34% reported dizziness. These were no significant differences between groups in the percentage of patients reporting side effects (6% in the M1 + CT and 7% in the DLPFC + CT). These side effects were mild and did not require any changes to the treatment protocol.

Primary outcome: fatigue

Fatigue was evaluated with the Fatigue Severity Scale (FSS), the Modified Fatigue Intensity Scale (MFIS) and with an Ecological Momentary Assessment.

At baseline, all patients had a diagnosis of fatigue (MFIS ≥ 38). From baseline to post-treatment, 25% of patients shifted from a fatigue diagnosis to no fatigue diagnosis (MFIS < 38). There were no significant differences between groups in the percentage of patients that changed diagnosis (29.03% for the M1 + CT, and 25% for the DLPFC + CT, p = 0.718). At follow-up, 14.7% of patients remained with no fatigue diagnosis (MFIS < 38), and there were no significant differences between groups (19.35% for M1 + CT and 12.5% for the DLPFC + CT, p = 0.457).

Physical fatigue measured through the FSS showed a significant main effect of time (p_bonf < 0.001), and no significant time-per-group interaction was found. Post hoc comparisons revealed a significant time effect between baseline and post-treatment (mean difference of 5.984, p_bonf < 0.001) and between baseline and follow-up (mean difference of 6.016, p_bonf < 0.001), but not between post-treatment and follow-up. That is, both groups improved in physical fatigue measured with the FSS and these improvements were maintained over one month follow-up (Fig. 1A, Table 2). Mean scores are shown in Supplementary Table S1.

[See PDF for image]

Fig. 1

Fatigue and clinical changes after treatment and follow-up. DLPFC, dorsolateral prefrontal cortex; M1, primary motor cortex; T0, pre-treatment; T1, post-treatment; T2, 1 month follow-up.

Table 2. Repeated measures ANOVA and post-hoc analyses in fatigue and other clinical aspects.

Outcome variable		Within-Subjects effects		Between-Subjects Effect	Post-hoc analysis (Time)
		Time	Time x Group interaction	Group	T0 vs T1	T0 vs T2	T1 vs T2
FSS	F or T (pbonf-value)	18.904 (< .001)	1.784 (0.172)	0.508 (0.479)	5.311 (< .001)	5.339 (< .001)	0.029 (0.977)
MFIS- total	F or T (pbonf-value)	39.284 (< .001)	0.535 (0.587)	0.535 (0.467)	8.138 (< .001)	7.112 (< .001)	-1.026 (0.307)
MFIS- physical	F or T (pbonf-value)	27.703 (< .001)	1.467 (0.235)	1.360 (0.248)	6.785 (< .001)	6.043 (< .001)	-0.742 (0.459)
MFIS- cognitive	F or T (pbonf-value)	37.020 (< .001)	0.735 (0.482)	0.033 (0.856)	7.472 (< .001)	7.432 (< .001)	-0.040 (0.968)
BDI-II	F or T (pbonf-value)	10.284 (< .001)	0.198 (0.820)	0.633 (0.429)	4.431 (< .001)	3.051 (0.006)	-1.380 (0.170)
PSQI	F or T (pbonf-value)	9.886 (< .001)	3.127 (0.047)	1.188 (0.280)	4.201 (< .001)	3.363 (0.002)	-0.838 (0.404)
STAI-state	F or T (pbonf-value)	3.304 (0.040)	2.004 (0.139)	0.009 (0.923)	2.523 (0.040)	1.725 (0.174)	-0.787 (0.433)
BPI-intensity	F or T (pbonf-value)	4.592 (0.012)	0.767 (0.467)	0.087 (0.770)	2.634 (0.029)	2.615 (0.029)	-0.020 (0.984)
BPI-functionality	F or T (pbonf-value)	8.286 (< .001)	0.626 (0.536)	0.104 (0.748)	4.056 (< .001)	1.731 (0.086)	-2.325 (0.043)
EuroQoL-5D	F or T (pbonf-value)	6.821 (0.002)	0.274 (0.761)	0.562 (0.456)	2.708 (0.016)	3.529 (0.002)	0.821 (0.413)

p_bonf, p value corrected by Holm–Bonferroni; FSS, Fatigue Severity Scale; MFIS, Modified Fatigue Impact Scale ; BDI, Beck Depression Inventory; PSQI, Pittsburgh Sleep Quality Index; STAI, State-Trait Anxiety Inventory; BPI, Brief Pain Inventory; EuroQoL-5D, Quality-of-Life scale.

For all MFIS subscales (MFIS Total, physical and cognitive), there was a significant main effect of time (p_bonf < 0.001), but no significant time-by-group interaction. Similarly, for all MFIS subscales, post-hoc analysis revealed significant improvement from baseline to post-treatment (all p_bonf < 0.001), and from baseline to follow-up (all p_bonf < 0.001), with no significant changes between post-treatment and follow-up (Fig. 1A, Table 2). Therefore, these results suggest that both treatments improved physical and cognitive fatigue measured with the MFIS and these improvements were maintained over time (Fig. 1A, Table 2). Mean scores are shown in Supplementary Table S1.

Ecological momentary assessment of physical and cognitive fatigue was also recorded (Fig. 2). Figure 2A shows the daily evolution of fatigue. Daily physical fatigue changes revealed a significant time-per-group effect between baseline and post-treatment (F = 4.099; p_bonf = 0.050) showing the M1 + CT (t = 4.672; p_bonf < 0.001) significant changes compared to the DLPFC + CT (t = 1.983; p = 0.208). A significant effect of time was also found (F = 22.599; p_bonf < 0.001). Daily cognitive fatigue changes revealed a significant effect of time between baseline and post-treatment (F = 14.294; p_bonf < 0.001) mostly influenced by the M1 + CT (t = 3.501; p_bonf = 0.008) rather than the DLPFC + CT (t = 1.803; p = 0.319), but no significant time-per-group effect was found. Figure 2B shows the percentage of daily change in physical or cognitive fatigue compared to the median fatigue score of the pre-treatment period.

[See PDF for image]

Fig. 2

Ecological Momentary Assessment of Fatigue. (A) Daily fatigue representation from 7 days before treatment to 30 days of follow-up. *reduction of 1 point in the Likert scale compared to the median value of the pre-treatment period. (B) Daily percentage of fatigue change compared to the median of the pre-treatment.

Compared to the median physical fatigue score of the pre-treatment period (M1 + CT median = 8; DLPFC + CT median = 7), the M1 + CT showed a reduction of 1 point in the Likert scale on the 10^th day of treatment, while the DLPFC + CT revealed a reduction of 1 point on day the 15^th day of treatment. Daily cognitive fatigue was also recorded. Compared to the median cognitive fatigue score of the pre-treatment period (M1 + CT median = 7; DLPFC + CT median = 7), patients from the M1 + CT revealed a 1-point reduction on the 7th day of treatment, while the DLPFC + CT showed a 1-point reduction on the 9th day of treatment (Fig. 2A).

Cognitive outcomes

Both groups significantly improved cognition, showing a significant effect of time in Stroop Word (p_bonf < 0.001), Stroop Color (p_bonf < 0.004), and Stroop Word-Color (p_bonf < 0.001), Symbol Digit Modality Test (p_bonf < 0.001), Digit Span Backwards (p_bonf = 0.011), Paced Auditory Serial Addition Test(p_bonf < 0.001), N-back (p_bonf = 0.005), and Free and Cued Selective Reminding Task (p_bonf < 0.001). No significant time-per-group interaction was found in any cognitive test (Fig. 3, Table 3).

[See PDF for image]

Fig. 3

Cognitive changes after treatment and follow-up. DLPFC, dorsolateral prefrontal cortex; M1, primary motor cortex; T0, pre-treatment; T1, post-treatment; T2, 1 month follow-up.

Table 3. Repeated measures ANOVA and post-hoc analyses in cognition.

Outcome variable		Within-subjects effects		Between-subjects effect	Post-hoc analysis (Time)
		Time	Time × Group interaction	Group	T0 vs T1	T0 vs T2	T1 vs T2
Stroop Word	F or T (pbonf-value)	14.593 (< .001)	0.623 (0.538)	3.694 (0.059)	− 3.758 (< .001)	− 5.240 (< .001)	− 1.483 (0.141)
Stroop Color	F or T (pbonf-value)	5.918 (0.004)	0.329 (0.720)	4.386 (0.040)	− 2.759 (0.013)	− 3.159 (0.006)	− 0.400 (0.690)
Stroop Word-Color	F or T (pbonf-value)	22.596 (< .001)	1.028 (0.361)	3.129 (0.082)	− 5.747 (< .001)	− 5.894 (< .001)	− 0.146 (0.884)
SDMT	F or T (pbonf-value)	20.107 (< .001)	0.263 (0.769)	1.558 (0.217)	3.648 (< .001)	− 6.316 (< .001)	− 2.668 (< .009)
Digit Span Backwards	F or T (pbonf-value)	4.709 (0.011)	1.528 (0.221)	1.750 (0.191)	− 2.087 (0.078)	− 2.992 (0.010)	− 0.905 (0.367)
PASAT	F or T (pbonf-value)	21.877 (< .001)	1.441 (0.241)	14.406 (< .001)	− 5.349 (< .001)	− 6.044 (< .001)	− 0.695 (0.488)
N-Back	F or T (pbonf-value)	5.472 (0.005)	2.592 (0.079)	1.339 (0.252)	− 3.254 (0.004)	− 2.144 (0.068)	1.110 (0.269)
FCSRT Free Trial 1	F or T (pbonf-value)	73.529 (< .001)	0.472 (0.625)	0.398 (0.531)	− 5.498 (< .001)	− 12.110 (< .001)	− 6.612 (< .001)
FCSRT Free Recall	F or T (pbonf-value)	55.480 (< .001)	2.335 (0.101)	1.907 (0.172)	− 5.177 (< .001)	− 10.533 (< .001)	− 5.356 (< .001)
FCSRT Total Recall	F or T (pbonf-value)	13.044 (< .001)	0.656 (0.521)	3.314 (0.074)	− 2.601 (0.021)	− 5.107 (< .001)	− 2.506 (0.021)
FCSRT Delay Free Recall	F or T (pbonf-value)	19.956 (< .001)	2.935 (0.057)	1.052 (0.309)	− 3.637 (< .001)	− 6.292 (< .001)	− 2.655 (0.009)
FCSRT Delay Recall	F or T (pbonf-value)	8.235 (< .001)	0.296 (0.745)	0.251 (0.618)	− 2.760 (0.013)	− 3.957 (< .001)	− 1.197 (0.234)
FLEI	F or T (pbonf-value)	22.595 (< .001)	2.511 (0.091)	0.512 (0.477)	5.634 (< .001)	5.993 (< .001)	0.360 (0.720)

p_bonf-, p value corrected by Holm–Bonferroni; FCSRT, Free and Cued Selective Remining Task; FLEI, Subjective Cognitive Performance Questionnaire; PASAT, Paced Auditory Serial Addition Test; SDMT, Symbol Digit mModality Test.

Post-hoc analyses revealed significant differences between baseline and post-treatment and between baseline and follow-up. That is, improvements in cognition were found after treatment, and these improvements were maintained after 1 month follow-up. Mean scores and mean change scores are shown in Supplementary Table S2. Interestingly, Symbol Digit Modality Test and Free and Cued Selective Reminding Task also revealed significant improvements from post-treatment to follow-up (p_bonf < 0.01), revealing that improvements may persist after treatment.

Subjective cognitive complaints were also significantly improved over time in both groups (p_bonf < 0.001), and no significant time-per-group effect was found. Post hoc analyses revealed significant differences between baseline and post-treatment (mean difference of 0.532, p_bonf = 0.001) and between baseline and follow-up (mean difference of 0.694, p_bonf < 0.001), but not between post-treatment and follow-up (Fig. 3, Table 3).

Secondary clinical outcomes

Both groups showed significant time effects for depression measured with the Beck Depression Inventory II (BDI-II) (p_bonf < 0.001) and anxiety measured with the State-Trait Anxiety Inventory (STAI) (p = 0.040) (Fig. 1B, Table 2). Post hoc analyses revealed significant improvements in depression from baseline to post-treatment (mean difference = 3.677, p_bonf < 0.001) and to follow-up (mean difference = 2.532, p_bonf = 0.006), but not between post-treatment and follow-up. For anxiety, significant changes were only observed from baseline to post-treatment (mean difference = 4.190, p = 0.040).

Concerning depression, at baseline, 44.4% of the patients had depression (BDI-II ≥ 19). From baseline to post-treatment, 12.7% of patients shifted from a depression to no depression (BDI-II < 18). There were no significant differences between groups in the percentage of patients that changed diagnosis (19.4% for the M1 + CT, and 6.5% for the DLPFC + CT, p = 0.051), however, paired-t-test showed that DLPFC + CT group showed a significant improvement (p < 0.001) compared to the M1 + CT group (p = 0.052). At follow-up, 61.9% of patients remained with no depression diagnosis (BDI-II < 18), and there were no significant differences between groups in the percentage of patients that changed diagnosis compared to baseline (19.4% for M1 + CT and 6.5% for the DLPFC + CT, p = 0.096).

Focusing on anxiety, at baseline, 25% of the patients had anxiety (state) (STAI ≥ 40). From baseline to post-treatment, 15.6% of patients shifted from anxiety to no anxiety (STAI < 39). There were no significant differences between groups in the percentage of patients that changed diagnosis (16.1% for the M1 + CT, and 15.6% for the DLPFC + CT, p = 0.853), however, paired-t-test showed that M1 + CT group showed a significant improvement (p = 0.007) compared to the DLPFC + CT group (p = 0.476). At follow-up, 79.4% of patients remained with no anxiety diagnosis (STAI < 39), and there were no significant differences between groups in the percentage of patients that changed diagnosis compared to baseline (16.1% for M1 + CT and 15.6% for the DLPFC + CT, p = 0.938).

Regarding other clinical aspects, both groups also showed improvements in pain (pain-Intensity: p = 0.012; pain-Functionality: p_bonf < 0.001). Pain-Intensity and pain-Functionality improved significantly from baseline to post-treatment (mean difference in pain-Intensity: 2.159, p_bonf = 0.029; mean difference in pain-Functionality (mean difference: 8.694, p_bonf < 0.001) with a sustained improvement in pain-Intensity at follow-up (mean difference: 2.143, p_bonf = 0.029), and late improvement in pain-Functionality between post-treatment and follow-up (mean difference: 4.984, p_bonf = 0.043) (Fig. 1B, Table 2).

Sleep quality also improved over time (p_bonf < 0.001), with a significant time-by-group interaction (p = 0.047), indicating greater improvements in the M1 + CT group (mean difference = 2.800, p_bonf =  < 0.001), with no significant changes in the DLPFC + CT group. Mean scores are detailed in Supplementary Table S1.

Quality of life also improved over time in both groups (p_bonf = 0.002), with no significant time by group interaction. Significant differences were found from baseline to post-treatment (mean difference = 0.532, p = 0.016) and follow-up (mean difference = 0.694, p_bonf = 0.002), but not between post-treatment and follow-up (Fig. 1B, Table 2).

Electric field simulation

The simulation of the electric field for the M1 + CT and DLPFC + CT is shown in Fig. 4. The mean field strength for the M1 + CT was of 0.533 ± 0.673 V/m while for the DLPFC + CT was of 0.415 ± 0.533 V/m showing significant differences between groups (t = 4.485; p_bonf < 0.001). The field focality for the M1 + CT was of 4.560 ± 3.328 cm³, while for the DLPFC + CT was of 4.984 ± 3.025 cm³ showing no significant differences between groups (t = 0.332; p = 0.734).

[See PDF for image]

Fig. 4

Electric field calculations. Visual representation of the electric field calculations performed in SimNIBS. Values are shown in V/m. First row shows mean of the normal field component simulations. Second row shows standard deviation of the normal field component simulations. A, anterior; P, posterior; S, superior; I, inferior; R, right; L, left.

Predictors of fatigue and cognition improvements

Regression analyses were performed to evaluate predictors of fatigue and cognitive improvement. The change in fatigue was assessed with the change in FSS as the primary objective of the study, and the change in cognition was assessed with a composite score including the change in Stroop Word, Stroop Word-Color, and Symbol Digit Modality Test, due to the relevance of these cognitive domains in post-COVID, in which the cognitive training was focused.

The change in depression showed a significant effect over the change in fatigue (F(2.91) = 8.52; p = 0.005; R2 = 0.10), explaining 10% of the change. Additionally, the change in Pain-Functionality revealed a significant effect over the change in fatigue (F(3.66) = 13.39; p = 0.001; R2 = 0.167), explaining 16% of the change.

Similarly, the change in depression showed a significant effect over the change in cognition (F(− 2.13) = 4.53; p = 0.037; R2 = 0.054), explaining 5.4% of the change. Finally, the change in pain-Functionality also revealed a significant effect over the change in cognition (F(− 3.74) = 14.05; p =  < 0.001; R2 = 0.174), explaining 17.4% of the change.

Regarding sleep, pain-Intensity or anxiety, no significant effects were found.

Discussion

The present study investigated the effectiveness of combining tDCS with CT to improve fatigue and cognitive dysfunction in post-COVID patients. Our previous work¹⁷ demonstrated that anodal tDCS applied over DLPFC significantly reduced physical fatigue compared to sham stimulation, although no significant effects were observed for other variables such as cognitive fatigue, cognition, mood or quality of life. Based on these previous findings, this study designed and evaluated the efficacy of a more comprehensive treatment aiming to improve not only fatigue but also cognition. Therefore, we conducted a randomized parallel trial to compare the effects of stimulation targeting the M1 with those of stimulation targeting the DLPFC, both paired with the same CT program. To date, very few studies have been carried out comparing two treatment targets for fatigue^26,27. Our findings suggested that the combination of tDCS and CT is safe, feasible and well-tolerated in patients with post-COVID. The sample of the present study was representative of post-COVID patients, as the majority were women (80%), which is consistent with findings from other studies that report a predominance of female participants, typically around 70–75%^{28, 29–30}. Additionally, the participants from the study were mostly middle-aged, which also aligns with the demographic profile commonly described in the post-COVID literature^31,32.

As we mentioned before, a proportion of individuals develop post-COVID, with only a subset achieving fully recovering^2,3. Fatigue is a primary and disabling symptom in post-COVID patients, underscoring the critical need for effective treatments. Regarding the primary objective of the study, which was to evaluate the effects over physical fatigue measured through the FSS, results revealed significant improvements in both groups, with the M1 + CT showing greater effects. Similar results were found with the MFIS scale. To assess whether these improvements translated into daily life, an ecological fatigue assessment was conducted, as it is more sensitive to daily fluctuations in fatigue and better captures short-term changes³³. Worth to highlight, M1 + CT group revealed a significant improvement in ecological fatigue assessment (both physical and cognitive) compared to DLPFC + CT group. These findings may suggest that the FSS, which is a brief scale of 9 items, and the MFIS scale, may not be too accurate to capture the specific changes in fatigue in post-COVID. Indeed, previous studies demonstrated the need of disease-specific scales in post-COVID to better capture fatigue symptoms^{34, 35–36}. The superior effects in fatigue and sleep quality observed in M1 + CT group may be partially explained by the higher electric field strength and broader cortical engagement compared to DLPFC stimulation³⁷. The anatomical location of M1 may facilitate broader modulation of brain regions implicated in fatigue, due to its proximity to sensorimotor, interoceptive and corticospinal networks³⁸, which are also involved in perception and arousal regulation—key processes associated with fatigue symptoms³⁹. Additionally, 25% of patients no longer met the clinical criteria for fatigue at post-treatment, indicating the intervention’s efficacy despite the fluctuating nature of fatigue⁴⁰.

Regarding cognition, the study also found improvements in various cognitive functions, including attention, working memory, processing speed, executive functions, and memory in both stimulation groups. Patients reported fewer cognitive complaints at post-treatment. The present results highlight the effectiveness of CT in improving cognition in post-COVID patients, in contrast to previous randomized controlled trials that used tDCS as a standalone treatment and found no significant improvements in cognition⁴¹. Indeed, a previous study on post-COVID patients supports these findings, showing cognitive gains in attention, memory, and reasoning following personalized CT⁴².

Beyond improvements in fatigue and cognitive function, the interventions also demonstrated beneficial effects on depression and anxiety symptoms. Consistent with previous clinical trials, stimulation of DLPFC remains a widely used target in neuromodulation studies aimed at alleviating depressive symptoms^27,41. Prior studies have identified the DLPFC as a key region in emotional regulation, with both structural and functional brain alterations commonly associated with depressive states⁴³. Moreover, emerging evidence suggests that stimulation of the primary motor cortex (M1) may contribute to the reduction of anxiety symptoms, potentially due to its anatomical and functional connectivity with sensory processing pathways⁴⁴.

Furthermore, focusing on clinical aspects, patients in the M1 + CT experienced significant improvements in sleep quality compared to DLPFC + CT group, which were maintained during follow-up. All these clinical benefits translated into enhanced quality of life, a key goal of treatment. In addition to reducing fatigue and enhancing cognitive function, both stimulation groups reported improvements in depression, anxiety, and pain. These are common symptoms in post-COVID patients, and their amelioration points to the broader therapeutic potential of the tDCS⁴⁵.

At one month follow-up, both groups maintained their gains in fatigue reduction and cognition, and quality of life. We hypothesized that these cognitive gains after one month may not be attributed to learning effects from repeated assessments, because parallel versions of the memory test were used, but to possible improvements in attention and working memory, which may have cascaded into other cognitive domains. In fact, attention, is a transversal cognitive domain; thus, the effects of its training have the potential to generalize to other cognitive domains. Moreover, attention, working memory and episodic memory are three interrelated cognitive domains^46,47. Nevertheless, some degree of practice effect cannot be completely ruled out; future studies should consider incorporating alternative versions of all tasks when available.

Few studies have explored the use of tDCS for fatigue in post-COVID patients. One prior study found significant improvements in cognitive fatigue when primary motor cortex stimulation was paired with physical training, although it did not observe significant effects on physical fatigue¹⁸. However, another study focusing on M1 stimulation did not report any significant outcomes⁴⁸. A pervious study from the current research group found that stimulation over the DLPFC improved physical fatigue but not cognitive fatigue or cognition¹⁷. These discrepancies may be explained by the shorter treatment durations in previous studies, as well as the absence of paired interventions, like cognitive or physical training.

Focusing on the predictors of fatigue and cognitive improvement, results revealed that the change in depression and pain-Functionality had low but significant effects over the change in fatigue and cognition. These findings suggest that changes in depression and pain were slightly associated with changes in fatigue and cognition.

Previous studies in post-COVID support the association between cognition and clinical aspects association⁴⁹. In other pathologies, depression has been shown to influence cognitive rehabilitation⁵⁰, and pain has been linked with cognitive performance⁵¹. Depression symptoms have been linked to increased fatigue and reduced cognitive performance in neurological populations⁵², while pain contributes to fatigue severity in chronic pain conditions such as fibromyalgia⁵³.

Although changes in depression and pain-Functionality reflected a significant effect on fatigue and cognition, the proportion of explained variance is relatively low. This suggests that the treatment had a direct effect on fatigue and cognition, beyond its impact on emotional symptoms. A similar pattern has been observed in other conditions, such as depression, where non-invasive brain stimulation over the DLPFC has a primary effect on depressive symptoms, but also produces secondary improvements in areas such as cognition, anxiety, and others⁵⁴. That is, improvements after training may facilitate secondary benefits in other aspects not treated directly, but interrelated⁵⁴.

Overall, these findings suggest the potential of brain stimulation paired with CT to improve fatigue and cognitive dysfunction in post-COVID patients. One of the main limitations of the study is the absence of a control group. Previous studies have already demonstrated the efficacy of the tDCS compared to placebo on improving fatigue and cognition in post-COVID patients^17,18 as well as in other neurological populations such as multiple sclerosis⁵⁵ or stroke⁵⁶ . Therefore, in the present study, we aimed to design and evaluate the efficacy of a more comprehensive treatment (tDCS + cognitive training) and compare the effectiveness of two targets that previously demonstrated efficacy against placebo in post-COVID (left DLPFC and left M1). However, the current design does not allow us to disentangle the individual effects of each intervention, as well as to fully exclude potential placebo effects. This limits the interpretability of the observed improvement in fatigue and cognition. Future trials should include a sham-controlled condition and single-treatment arms, to better determine the specific individual and combined contributions of each treatment. Furthermore, longitudinal studies with comparable or extended follow-up periods than the present study suggest that post-COVID symptoms generally persist over time, underscoring the necessity for interventions aimed at mitigating these enduring symptoms^10,11.

Additionally, the follow-up period was limited to one month. While improvements in fatigue, cognition and quality of life were maintained during one month, it remains unclear whether these effects are transient or enduring. Therefore, long-term follow-up assessments are needed to determine the durability of the benefits observed in this pathology, and whether booster sessions may be necessary to sustain improvements. Moreover, while the study evaluated the main cognitive domains affected in post-COVID patients, future research should broaden the scope to include areas like language, visuospatial abilities, and social cognition, which are also relevant to the cognitive profile of post-COVID individuals. Another limitation is the relatively small sample size, although it is comparable to previous studies using a multisession tDCS protocol^18,55 it may have limited the generalizability of the findings. Nevertheless, the fact that most participants completed the protocol highlights not only the feasibility of the intervention but also the high level of adherence, which is a key factor when considering the potential for its real clinical application. Furthermore, although both participants and outcome assessors were rigorously blinded throughout the study, the effectiveness of blinding was not objectively evaluated. Future randomized controlled trials should include a blinding assessment in their protocols.

In conclusion, combining tDCS with CT is a feasible and safe treatment for post-COVID patients, leading to significant improvements in fatigue, cognitive function, and overall quality of life. While both stimulation targets were beneficial, the primary motor cortex target appeared slightly more effective in reducing fatigue and improving sleep quality. Given these promising findings, further randomized controlled trials are recommended to confirm the treatment’s efficacy and to optimize protocols for broader application in post-COVID.

Methods

Design and ethics statement

This randomized, parallel trial compared stimulation over two tDCS targets combined with cognitive intervention. Assessments were conducted at three time points: baseline, post-treatment, and a one-month follow-up. Figure 5 provides a summary of the study flow. The study was conducted at the Department of Neurology of Hospital Clinico San Carlos, in Madrid, from 1^st March to 30^th June 2023, and was approved by the local Ethics Committee (CEIm Hospital Clínico San Carlos) (22/728-P_EC). Participants were recruited through two channels: referrals from the Neurology Department at Hospital Clinico San Carlos and collaboration with local patients’ association. Participants provided written informed consent before enrollment, and the protocol was registered on ClinicalTrials.gov (NCT05753202).

[See PDF for image]

Fig. 5

CONSORT Flow Diagram. CONSORT, consolidated standards of reporting trials.

Participants

Post-COVID patients were enrolled in the study if they had: (1) Diagnosis of COVID-19 infection confirmed by RT-PCR at least six months before the inclusion in the study; (2) Diagnosis of post-COVID condition according to the WHO criteria¹; (3) Symptoms of fatigue reported by the patient. Fatigue symptoms should be temporally associated with SARS-CoV-2 infection and severe enough to interfere with daily activities (MFIS ≥ 38); (4) Native proficiency in written Spanish; 4) Voluntary participation and signed the informed consent; (5) Participant has not been previously diagnosed with a neurological disorder or other systemic disease that may affect fatigue; (6) Participant has not been previously diagnosed with, nor is undergoing treatment for an active psychiatric disorder that may affect fatigue or cognition (e.g., schizophrenia). Exclusion criteria are specified in supplementary materials.

Randomization and blinding

Seventy-one patients with post-COVID were invited to participate. Four declined the invitation therefore, 67 participants were recruited. After the baseline assessment (T0), patients were randomly allocated at 1:1 ratio to two groups using www.studyrandomizer.com. Simple randomization was employed without stratification. Both patients and outcome assessors were blinded to group allocation. The researcher administering the tDCS sessions was also responsible for patient enrollment and had no access to assessment results. Group allocation remained concealed until the last participant completed the final study assessment.

Sample size estimation

The primary endpoint of the present study was fatigue, assessed using the Fatigue Severity Scale (FSS). Previous studies have identified a 0.88-point change in the FSS as clinically significant⁵⁷. Assuming Type I error of 0.05, a power of 80%, and a minimum detectable change between groups of 1 point on the FSS, the required sample size was calculated to be 34 subjects (17 per group). Considering an anticipated dropout rate of 20%, the minimum sample size required was set at 42 subjects (21 per group).

Fatigue assessment

The primary endpoint of the study was the change in physical fatigue assessed with the FSS. The FSS⁵⁸ is one of the most frequently used scales for measuring fatigue. It evaluates fatigue as a unidimensional construct and consists of 9 items, each rated in a seven-point Likert scale ranging from 1 “strongly disagree” to 7 “strongly agree”, with higher scores indicating greater fatigue.

Second, the MFIS is a multidimensional scale that evaluates physical, cognitive and psychosocial fatigue⁵⁹. The MFIS questionnaire includes 21 items, scored from 0 “never” to 4 “almost always”. This scale includes a total score (MFIS total score) and three subscales: (1) a Cognitive subscale (MFIS Cognitive subscale), that ranges from 0 to 40; (2) a Physical subscale (MFIS Physical subscale) ranging from 0 to 36; and (3) a Psychosocial subscale, ranging from 0 to 8. Higher scores indicate greater fatigue. Fatigue diagnosis is established with the cut-off of MFIS total score ≥ 38⁶⁰.

Ecological momentary assessment of fatigue

An ecological momentary assessment was conducted to evaluate physical and cognitive fatigue symptoms. The aim was to obtain daily data on patients’ fatigue while they were engaged in their daily routines. Fatigue levels were recorded daily from 10 days before the first treatment session until one month after post-treatment (T2). Daily physical and cognitive fatigue were recorded using a Likert scale from 0 to 10, with higher scores indicating greater fatigue. Patients completed the survey via email.

Neuropsychological and clinical assessment

Post-COVID patients underwent a comprehensive neuropsychological and clinical evaluation. A trained neuropsychologist administered the cognitive protocol including attention, working memory, processing speed, executive functions, and memory. Specific cognitive tests are specified in Supplementary materials. Moreover, a subjective cognitive performance questionnaire was also administered. Clinical assessment also included the assessment of anxiety with the STAI⁶¹ (cut-off score STAI-state ≥ 40 was considered clinically significant anxiety), depression with BDI-II⁶² (cut-off score BDI-II ≥ 19 was considered moderate or severe depression⁶³) sleep quality with the Pittsburgh Sleep Quality Index (PSQI)⁶⁴, pain was assessed with the Brief Pain Inventory⁶⁵ and quality of life was evaluated with the Quality-of-life scale (EuroQol-5D)⁶⁶.

Interventions

Both treatment groups underwent tDCS combined with the same CT sessions. Each treatment consisted of 15 sessions, 1 session/day, 20 min/session, Monday to Friday. Ideally, patients attended five sessions per week, with a minimum requirement of three. All participants completed the 15 sessions. One treatment arm received active tDCS with the anode placed over the left M1 while undergoing CT (M1 + CT). The second treatment arm received repeated anodal stimulation over the left DLPFC while undergoing the same CT (DLPFC + CT). Both left DLPFC and left M1 targets were chosen due to previous significant results in post-COVID fatigue^17,18 and in other pathologies¹⁴.

tDCS protocol treatment

Stimulation was administered using a Nurostym tES device (NeuroDevice Group S.A.; Poland). Patients received active stimulation via sponge-covered electrodes soaked in saline solution. In some cases, electroconductive gel was applied to reduce impedance. The anode (electrode size 5 × 7 cm) was positioned over the left M1 (C3, according to the international 10–20 system) or the left DLPFC (F3). In both groups, the cathode (size 5 × 7 cm) was placed over the contralateral supraorbital region (FP2). The anode electrode was precisely positioned using neuronavigation using a standard template and the BrainSight neuronavigation system (Rogue Research Inc.; Canada).

The device was programmed to deliver a constant current of 2 mA intensity for 20 min. At the beginning of each session, the current was gradually ramped up over the first 10 s until reaching 2 mA. At the end of the session, it was gradually ramped down until it reaching 0mA. Impedance values were continuously monitored during the stimulation and maintained below 10 kΩ.

Adaptive cognitive training

Both groups underwent adaptive CT during the 20 min tDCS sessions. Adaptive CT has demonstrated to be more effective in improving cognition than non-adaptive training⁶⁷. The training was conducted individually using a computerized program NeuronUP®⁶⁸, and targeted the cognitive domains most affected in patients with post-COVID: attention (first 5 days), working memory (next 5 days), and processing speed (last 5 days)⁶⁹. All participants began the exercises at the same difficulty level, which was then automatically adjusted based on their performance. The same cognitive training protocol was applied in both treatment arms.

Feasibility and safety

Feasibility was assessed based on the proportion of participants who completed the treatment. Safety assessments were conducted in accordance with Good Clinical Practice standards and current regulations. Adverse events were recorded at each visit, thoughout the treatment period, including their type and frequency. The study was conducted at the Hospital, ensuring that medical response and nursing support were available from both the research team and hospital staff if needed.

Electric field simulation

Magnetic resonance imaging data were available for 25 patients (13 from the M1 + CT and 12 from the DLPFC + CT) for electric field simulations. Images were acquired using a 3T GE Signa Architect scanner with a 48-channel head coil. High resolution 3D T1-weighted images were obtained using a sagittal MPRAGE sequence. Detailed magnetic resonance imaging acquisition parameters and electric field simulation analyses are provided in the supplementary materials.

Statistical analyses

Statistical analyses were conducted using IBM® SPSS Statistics 20.0 and JASP. The normality of the data distribution was evaluated with Shapiro–Wilk test. Baseline characteristics between groups were compared using the χ² test or t-tests, as appropriate. The effect of treatment and its longitudinal maintenance were evaluated using repeated-measures ANOVA, with time (baseline, post-treatment and follow-up) as a within-subjects factor, and group as a between-subjects factor. The ‘time-by-group’ interaction was estimated. Mauchly’s test was used to determine whether the assumption of sphericity was met, and Levene’s test was used to assess the homogeneity of variance. Post hoc analyses were performed for statistically significant main effects or interactions, with p-values adjusted for multiple comparisons according to the Holm–Bonferroni method (p_bonf). Finally, regression analyses were performed to evaluate predictors of fatigue and cognitive improvement.

Acknowledgements

The authors thank all the participants in this study, and specifically the long COVID patients’ association ‘Asociación Madrileña de COVID-19 Persistente’ (AMACOP) for their support and cooperation. The authors would also like to thank NeuronUP.

Author contributions

Conception: S.O.M., J.A.M.G., M.D.C.; Acquisition: S.O.M., C.D.A, L.G.M, C.C, L.F.R, A.M.G; Analysis: S.O.M., M.D.C.; Interpretation: all authors; Supervision: M.D.C., J.A.M.G., Writing – Original Draft Preparation: S.O.M.; Writing—Review and Editing and approved the submitted version: all authors.

Funding

Nominative Grant FIBHCSC 2020 COVID-19. Department of Health, Community of Madrid (JMG, JAMG). Instituto de Salud Carlos III through the project INT20/00079 and INT23/00017, co-funded by the European Regional Development Fund “A way to make Europe” (JAMG). Instituto de Salud Carlos III (ISCIII) through Sara Borrell postdoctoral fellowship Grant No. CD22/00043) and co-funded by the European Union (MDC). Fundación para el Conocimiento madri + d through the project G63-HEALTHSTARPLUS-HSP4 (JAMG, SOM).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval and informed consent

The present study was approved by the Research Ethics Committee from Hospital Clinico San Carlos (Approval number: 22/728-P_EC) and participants provided written informed consent prior to research participation. All research was performed in accordance with the Declaration of Helsinki.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Soriano, JB; Murthy, S; Marshall, JC; Relan, P; Diaz, JV. A clinical case definition of post-COVID-19 condition by a Delphi consensus. Lancet Infect. Dis.; 2022; 22, pp. e102-e107.1:CAS:528:DC%2BB38XislSrsg%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34951953]

2. Thaweethai, T et al. Development of a definition of postacute sequelae of SARS-CoV-2 infection. JAMA; 2023; 329, pp. 1934-1946.1:CAS:528:DC%2BB3sXht1KqtL%2FL [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37278994][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10214179]

3. Mateu, L et al. Determinants of the onset and prognosis of the post-COVID-19 condition: A 2-year prospective observational cohort study. Lancet Reg. Health-Europe; 2023; 33, 100724.

4. Premraj, L et al. Mid and long-term neurological and neuropsychiatric manifestations of post-COVID-19 syndrome: A meta-analysis. J. Neurol. Sci.; 2022; 434, 1:CAS:528:DC%2BB38XislGlsLg%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35121209][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8798975]120162.

5. Fernández-de-Las-Peñas, C et al. Prevalence of post-COVID-19 symptoms in hospitalized and non-hospitalized COVID-19 survivors: A systematic review and meta-analysis. Eur. J. Intern. Med.; 2021; 92, pp. 55-70. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34167876][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8206636]

6. Delgado-Alonso, C et al. Fatigue and cognitive dysfunction are associated with occupational status in post-COVID syndrome. Int. J. Environ. Res. Public Health; 2022; [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36293950][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9603617][DOI: https://dx.doi.org/10.3390/ijerph192013368]

7. Heine, J et al. Structural brain changes in patients with post-COVID fatigue: A prospective observational study. EClinicalMedicine; 2023; 58, 101874. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36873426][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9969172]

8. Douaud, G et al. SARS-CoV-2 is associated with changes in brain structure in UK Biobank. Nature; 2022; 604, pp. 697-707.2022Natur.604.697D1:CAS:528:DC%2BB38XhtVGnsLbM [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35255491][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9046077]

9. Díez-Cirarda, M et al. Multimodal neuroimaging in post-COVID syndrome and correlation with cognition. Brain; 2022; [DOI: https://dx.doi.org/10.1093/brain/awac384]

10. Taquet, M et al. Cognitive and psychiatric symptom trajectories 2–3 years after hospital admission for COVID-19: A longitudinal, prospective cohort study in the UK. Lancet Psychiatry; 2024; 11, pp. 696-708. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39096931]

11. Zhang, H et al. 3-year outcomes of discharged survivors of COVID-19 following the SARS-CoV-2 omicron (B.1.1.529) wave in 2022 in China: A longitudinal cohort study. Lancet Respir. Med.; 2024; 12, pp. 55-66. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38000376]

12. Markser, A. et al. Non-invasive brain stimulation for post-COVID-19 conditions: A systematic review. Am. J. Med. (2024).

13. Chase, HW; Boudewyn, MA; Carter, CS; Phillips, ML. Transcranial direct current stimulation: A roadmap for research, from mechanism of action to clinical implementation. Mol. Psychiatry; 2020; 25, pp. 397-407. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31455860]

14. Kan, RLD et al. Effects of non-invasive brain stimulation in multiple sclerosis: Systematic review and meta-analysis. Ther. Adv. Chronic Dis.; 2022; 13, 20406223211069200.

15. Fitzgibbon, BM et al. Evidence for the improvement of fatigue in fibromyalgia: A 4-week left dorsolateral prefrontal cortex repetitive transcranial magnetic stimulation randomized-controlled trial. Eur. J. Pain; 2018; 22, pp. 1255-1267.1:STN:280:DC%2BC1MnhsValsA%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29542208]

16. Pinto, ACPN et al. Transcranial direct current stimulation for fatigue in patients with Sjogren’s syndrome: A randomized, double-blind pilot study. Brain Stimul.: Basic Transl. Clin. Res. Neuromodul.; 2021; 14, pp. 141-151.

17. Oliver-Mas, S et al. Transcranial direct current stimulation for post-COVID fatigue: A randomized, double-blind, controlled pilot study. Brain Commun.; 2023; 5, fcad117. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37091591][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10116605]

18. Santana, K et al. Non-invasive brain stimulation for fatigue in post-acute sequelae of SARS-CoV-2 (PASC). Brain Stimul.; 2023; 16, pp. 100-107. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36693536][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9867562]

19. Martin, DM; Liu, R; Alonzo, A; Green, M; Loo, CK. Use of transcranial direct current stimulation (tDCS) to enhance cognitive training: Effect of timing of stimulation. Exp. Brain Res.; 2014; 232, pp. 3345-3351. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24992897]

20. Kelly, ME et al. The impact of cognitive training and mental stimulation on cognitive and everyday functioning of healthy older adults: A systematic review and meta-analysis. Ageing Res. Rev.; 2014; 15, pp. 28-43.2014ligh.book...K [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24607830]

21. Lampit, A et al. Computerized cognitive training in multiple sclerosis: A systematic review and meta-analysis. Neurorehabil. Neural Repair; 2019; 33, pp. 695-706. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31328637]

22. Díez-Cirarda, M et al. Long-term effects of cognitive rehabilitation on brain, functional outcome and cognition in Parkinson’s disease. Eur. J. Neurol.; 2018; 25, pp. 5-12. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28940855]

23. Ke, Y; Liu, S; Chen, L; Wang, X; Ming, D. Lasting enhancements in neural efficiency by multi-session transcranial direct current stimulation during working memory training. NPJ Sci. Learn.; 2023; 8, 48. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37919371][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10622507]

24. Ruf, SP; Fallgatter, AJ; Plewnia, C. Augmentation of working memory training by transcranial direct current stimulation (tDCS). Sci. Rep.; 2017; 7, 876.2017NatSR..7.876R [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28432349][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5430723]

25. Antonenko, D et al. Microstructural and functional plasticity following repeated brain stimulation during cognitive training in older adults. Nat. Commun.; 2023; 14, 3184.2023NatCo.14.3184A1:CAS:528:DC%2BB3sXhtFGkt7zN [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37268628][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10238397]

26. Lefaucheur, J-P et al. The treatment of fatigue by non-invasive brain stimulation. Neurophysiol. Clin.; 2017; 47, pp. 173-184. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28410876]

27. Hsu, W-Y; Cheng, C-H; Zanto, TP; Gazzaley, A; Bove, RM. Effects of transcranial direct current stimulation on cognition, mood, pain, and fatigue in multiple sclerosis: A systematic review and meta-analysis. Front. Neurol.; 2021; 12, 626113. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33763014][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7982804]

28. Sk Abd Razak, R et al. Post-COVID syndrome prevalence: A systematic review and meta-analysis. BMC Public Health; 2024; 24, 1785. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38965510][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11223303]

29. de Melo, JPB et al. Demographic, epidemiological and clinical profile of patients with post-COVID syndrome followed at a teaching hospital in Brazil. Braz. J. Infect. Dis.; 2025; 29, [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39978115][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11889545]104509.

30. Bai, F et al. Female gender is associated with long COVID syndrome: a prospective cohort study. Clin. Microbiol. Infect.; 2022; 28, pp. 611-e9.

31. Chen, C et al. Global prevalence of post-coronavirus disease 2019 (COVID-19) condition or long COVID: A meta-analysis and systematic review. J. Infect. Dis.; 2022; 226, pp. 1593-1607.1:CAS:528:DC%2BB3sXptFartbo%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35429399]

32. Morioka, S et al. Epidemiology of post-COVID conditions beyond 3 years and factors associated with their persistence longer than 2 years: A cross-sectional study. J. Infect. Chemother.; 2024; 30, pp. 734-740. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38350510]

33. Shiffman, S; Stone, AA; Hufford, MR. Ecological momentary assessment. Annu. Rev. Clin. Psychol.; 2007; 4, pp. 1-32.

34. Oliver-Mas, S et al. Differential fatigue profile in patients with post-COVID condition, fibromyalgia, and multiple sclerosis. J. Clin. Med.; 2025; 14, 952.1:CAS:528:DC%2BB2MXktVCju74%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39941623][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11818582]

35. Curran, SL; Beacham, AO; Andrykowski, MA. Ecological momentary assessment of fatigue following breast cancer treatment. J. Behav. Med.; 2004; 27, pp. 425-444. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15675633]

36. Kratz, AL; Murphy, SL; Braley, TJ. Ecological momentary assessment of pain, fatigue, depressive, and cognitive symptoms reveals significant daily variability in multiple sclerosis. Arch. Phys. Med. Rehabil.; 2017; 98, pp. 2142-2150. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28729168][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5660933]

37. Janssen, AM; Oostendorp, TF; Stegeman, DF. The coil orientation dependency of the electric field induced by TMS for M1 and other brain areas. J. Neuroeng. Rehabil.; 2015; 12, 47. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25981522][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4435642]

38. Silva-Batista, C; Arpan, I; Penteado Nucci, M; Horak, FB. Alterations in functional connectivity of interoceptive and cortico-subcortical areas in multiple sclerosis-related fatigue. Neuroscience; 2025; 568, pp. 355-363.1:CAS:528:DC%2BB2MXislers7o%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39709059]

39. Wyller, VB; Eriksen, HR; Malterud, K. Can sustained arousal explain the chronic fatigue syndrome?. Behav. Brain Funct.; 2009; 5, 10. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19236717][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2654901]

40. Buttery, S et al. Patient symptoms and experience following COVID-19: Results from a UK-wide survey. BMJ Open Respir. Res.; 2021; 8, [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34732518]e001075.

41. Meron, D; Hedger, N; Garner, M; Baldwin, DS. Transcranial direct current stimulation (tDCS) in the treatment of depression: Systematic review and meta-analysis of efficacy and tolerability. Neurosci. Biobehav. Rev.; 2015; 57, pp. 46-62. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26232699]

42. Duñabeitia, JA; Mera, F; Baro, O; Jadad-Garcia, T; Jadad, AR. Personalized computerized training for cognitive dysfunction after COVID-19: A before-and-after feasibility pilot study. Int. J. Environ. Res. Public Health; 2023; 20, 3100. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36833793][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9966004]

43. Zheng, EZ; Wong, NML; Yang, ASY; Lee, TMC. Evaluating the effects of tDCS on depressive and anxiety symptoms from a transdiagnostic perspective: A systematic review and meta-analysis of randomized controlled trials. Transl. Psychiatry; 2024; 14, 295. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39025832][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11258305]

44. Sankarasubramanian, V et al. Transcranial direct current stimulation targeting primary motor versus dorsolateral prefrontal cortices: Proof-of-concept study investigating functional connectivity of thalamocortical networks specific to sensory-affective information processing. Brain Connect.; 2017; 7, pp. 182-196. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28142257][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5399740]

45. Lefaucheur, J-P et al. Evidence-based guidelines on the therapeutic use of transcranial direct current stimulation (tDCS). Clin. Neurophysiol.; 2017; 128, pp. 56-92. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27866120]

46. Shipstead, Z; Lindsey, DRB; Marshall, RL; Engle, RW. The mechanisms of working memory capacity: Primary memory, secondary memory, and attention control. J. Mem. Lang.; 2014; 72, pp. 116-141.

47. Chun, MM; Turk-Browne, NB. Interactions between attention and memory. Curr. Opin. Neurobiol.; 2007; 17, pp. 177-184.1:CAS:528:DC%2BD2sXksFehtrY%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17379501]

48. Klírová, M et al. Transcranial direct current stimulation (tDCS) in the treatment of neuropsychiatric symptoms of long COVID. Sci. Rep.; 2024; 14, 2193.2024NatSR.14.2193K [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38272997][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10810850]

49. Delgado-Alonso, C. et al. Unraveling brain fog in the post-COVID syndrome: relationship between subjective cognitive complaints and cognitive function, fatigue, and neuropsychiatric symptoms. Eur. J. Neurol. in press (2023).

50. Quattropani, MC; Geraci, A; Lenzo, V; Chiaie, RD; Filastro, A. Post-stroke anxiety and depression: Relationships to cognitive rehabilitation outcome. Clin. Neuropsychiatry; 2018; 15, pp. 12-18.

51. Sjøgren, P; Thomsen, AB; Olsen, AK. Impaired neuropsychological performance in chronic nonmalignant pain patients receiving long-term oral opioid therapy. J. Pain Symptom Manag.; 2000; 19, pp. 100-108.

52. Golan, D et al. The impact of subjective cognitive fatigue and depression on cognitive function in patients with multiple sclerosis. Mult. Scler. J.; 2018; 24, pp. 196-204.

53. Vincent, A et al. Beyond pain in fibromyalgia: Insights into the symptom of fatigue. Arthritis Res. Ther.; 2013; 15, pp. 1-14.

54. Kan, RLD et al. Effects of repetitive transcranial magnetic stimulation of the left dorsolateral prefrontal cortex on symptom domains in neuropsychiatric disorders: A systematic review and cross-diagnostic meta-analysis. Lancet Psychiatry; 2023; 10, pp. 252-259. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36898403]

55. Ferrucci, R et al. Transcranial direct current stimulation (tDCS) for fatigue in multiple sclerosis. NeuroRehabilitation; 2014; 34, pp. 121-127. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24284464]

56. Dong, X-L et al. A randomized controlled trial to explore the efficacy and safety of transcranial direct current stimulation on patients with post-stroke fatigue. Medicine; 2021; 100, 1:CAS:528:DC%2BB3MXitlyls7bK [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34731132][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8519229]e27504.

57. Rooney, S; McFadyen, A; Wood, L; Moffat, F; Paul, L. Minimally important difference of the fatigue severity scale and modified fatigue impact scale in people with multiple sclerosis. Mult. Scler. Relat. Disord.; 2019; 35, pp. 158-163. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31400557]

58. Lerdal, A. Fatigue Severity Scale. in Encyclopedia of Quality of Life and Well-Being Research (ed. Maggino, F.) 1–5 (Springer International Publishing, 2020). https://doi.org/10.1007/978-3-319-69909-7_1018-3.

59. Larson, RD. Psychometric properties of the modified fatigue impact scale. Int. J. MS Care; 2013; 15, pp. 15-20. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24453758][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3883028]

60. Strober, LB et al. Tired of not knowing what that fatigue score means? Normative data of the Modified Fatigue Impact Scale (MFIS). Mult. Scler. Relat. Disord.; 2020; 46, 1:STN:280:DC%2BB3szlsVaisg%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33296974]102576.

61. Spielberger, C. D., Gonzalez-Reigosa, F., Martinez-Urrutia, A., Natalicio, L. F. S. & Natalicio, D. S. The state-trait anxiety inventory. Rev. Interam. Psicol.5, (2017).

62. Beck, AT; Steer, RA; Brown, GK. Beck depression inventory-II. San Antonio; 1996; 78, pp. 490-498.

63. Beck, AT; Steer, RA; Carbin, MG. Psychometric properties of the Beck depression inventory: Twenty-five years of evaluation. Clin. Psychol. Rev.; 1988; 8, pp. 77-100.

64. Buysse, DJ; Reynolds, CF; Monk, TH; Berman, SR; Kupfer, DJ. The Pittsburgh sleep quality index: A new instrument for psychiatric practice and research. Psychiatry Res.; 1989; 28, pp. 193-213.1:STN:280:DyaL1Mzhs1Wmsg%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/2748771]

65. Atkinson, TM et al. Using confirmatory factor analysis to evaluate construct validity of the brief pain inventory (BPI). J. Pain Symptom Manag.; 2011; 41, pp. 558-565.

66. Herdman, M; Badia, X; Berra, SE. EuroQol-5D: Una alternativa sencilla para la medición de la calidad de vida relacionada con la salud en atención primaria. Aten Primaria; 2001; 28, pp. 425-429.1:STN:280:DC%2BD3MrlsFCktg%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11602124][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7684037]

67. Pedullà, L et al. Adaptive vs. non-adaptive cognitive training by means of a personalized app: A randomized trial in people with multiple sclerosis. J. Neuroeng. Rehabil.; 2016; 13, 88. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27716336][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5050994]

68. NeuronUP, S. L. NeuronUP. Preprint at (2012).

69. Delgado-Alonso, C et al. Cognitive dysfunction associated with COVID-19: A comprehensive neuropsychological study. J. Psychiatr. Res.; 2022; 150, pp. 40-46. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35349797][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8943429]