Fibromyalgia (FM) comprises widespread chronic pain and concurs with significant emotional distress associated with functional disability for daily activities. The symptoms linger or recur for at least 3 months without other conditions explaining the pain (Treede et al., 2015; Dueñas et al., 2016). The symptoms’ severity scale of the American College of Rheumatology (ACR, 2016) diagnosis criteria included cognitive impairment as an element of core symptoms of FM (Montoro et al., 2015). Attention, perception, memory, executive functioning, and language abilities are essential components of cognition (Gellman and Rick Turner, 2013). The processing of cognitive components includes active decision-making, learning, and memory of past events (Hansen and Streltzer, 2005; Moriarty et al., 2011). This complex processing involves extensive cortical and subcortical neural circuitry responsible for perception, localization, processing, relaying, and pain modulation. Thus, the pain experience is modulated by affective-motivational and cognitive-evaluative components than being a purely sensory phenomenon (Tyng et al., 2017). Chronic pain syndromes, such as FM, have been linked to cognitive processing disturbance (Khera and Rangasamy, 2021).

There is evidence that pain and neurocognition have anatomical, biochemical, and molecular associations (Khera and Rangasamy, 2021). The frontal lobes control executive functions, particularly the orbitofrontal cortex, anterior cingulate cortex, and dorsolateral prefrontal cortex (DLPC) (Verdejo-García et al., 2009). The somatosensory cortex distinguishes between painful and non-painful sensations, whereas the medial thalamus and anterior cingulate cortex (ACC) record the stimuli as painful. The emotional component of pain perception and memory formation are both impacted by this encoding process, which is also linked to improved functional connectivity between the thalamus and the mPFC (Tseng et al., 2017). There is an overlap of brain structures involved in executive function and pain perception, and either cognitive impairment or chronic pain involves maladaptive neuroplasticity processes (Khera and Rangasamy, 2021). Higher executive functioning requires the ability to make emotional decisions (Tyng et al., 2017). They include executive function, learning, memory, sustained focus, processing speed, and psychomotor ability (Khera and Rangasamy, 2021). The cognitive impairment hinders interaction with the environment and generates difficulties with working memory (WM) (Miller et al., 2018). The WM is responsible for the temporary storage and manipulation of information necessary to perform complex tasks, such as language comprehension, learning, and reasoning (Cowan, 2014). It is essential to the adequate performance of complex behaviors. Hence, when it fails, so does the capacity to carry out daily living activities and the ability to elaborate pain confrontation strategies (D’Esposito and Postle, 2015). In FM, the core complaints related to cognitive impairment are mental confusion, concentration difficulties, and failing memory. This set of symptoms is often called “FibroFog” (Kravitz and Katz, 2015; Walitt et al., 2016; Bell et al., 2018). According to a recent study, FM patients performed less accurately on activities requiring split attention and attentional switching (Moore et al., 2019). Regardless of chronic pain impact on cognitive impairment in FM, it does not seem to correlate with other musculoskeletal or neuropathic pain (Grisart and Van der Linden, 2001; Verdejo-García et al., 2009).

Clinical and preclinical studies indicate a bidirectional link between cognition and chronic pain (Serrano et al., 2022). However, the targets of treatment of chronic pain comprise modulation of the central sensory processing either pain transmission [i.e., opioids and tricyclic antidepressants (TCAs) or neural excitability should be the therapeutic targets (i.e., opioids, anticonvulsants)]. However, multiple medicines are needed to treat FM symptoms; some might worsen cognitive impairment (i.e., opioids) (Ngian et al., 2011). Despite the modulation of the central sensory pain processing to be a treatment target, the pharmacological approaches might be ineffective in many patients (Schiltenwolf et al., 2014), and some of them can worsen cognitive performance, so the interest in non-pharmacological interventions. Among these interventions, transcranial direct current stimulation (tDCS) has demonstrated clinical benefits for complex chronic pain conditions, such as FM (Zortea et al., 2019). The main target to apply the anodal(a)-tDCS for pain is the primary motor cortex (M1), based on the rationale that it enhances the excitability of the sensory-discriminative networks (Zortea et al., 2019). Another potential target area to apply the a-tDCS is over DLPFC since it has been found to have beneficial effects on mood regulation, cognitive functions, and maladaptive emotional functioning (Dixon et al., 2017; Sankarasubramanian et al., 2017). Regarding the a-tDCS impact on FM, its use on the left-(l)-DLPFC revealed benefits on cognitive performance (Santos et al., 2018), and its use at home was effective in improving pain (Brietzke et al., 2020) and pain catastrophizing (Caumo et al., 2022).

The a-tDCS can modulate cortical and subcortical neural networks, inducing a top-down effect. Its effect on healthy controls (HC) demonstrates that a-tDCS over the l-DLPFC improved digit-span performance (Rottschy et al., 2012; Barbey et al., 2013). However, other studies found that it enhanced digit-span performance only if the stimulus had been paired with an online WM (n-back) task (Hill et al., 2019). Additionally, we showed that the alertness, orienting, and executive control attentional networks are all modulated by a single session of a-tDCS with 2 mA administered to the l-DLPFC in combination with a Go/No-go test (Silva et al., 2017). Besides, studies found that tDCS’s impact on pain and cognitive function is neuroplasticity state-dependent, as indexed by the brain-derived-neurotrophic factor (BDNF) (Santos et al., 2018; Brietzke et al., 2019; da Graca-Tarragó et al., 2019). In the same perspective, earlier studies found that serum BDNF is associated positively with the descending pain modulatory system malfunction (DPMS) (Caumo et al., 2016; Soldatelli et al., 2021) and that the BDNF likely mediates the a-tDCS effect in the improvement of DPMS (da Graca-Tarragó et al., 2019; Beltran Serrano et al., 2020). Hence, substantial evidence supports the critical role of BDNF in synaptic plasticity, learning, and memory (Kowianski et al., 2018), and a decrease of this neurotrophic factor in the hippocampus is related to the worst cognitive performance on memory tasks (Etnier et al., 2015). In this setting, it is reasonable to consider the BDNF as a neural plasticity marker involved in the tDCS effects, either on pain processing or cognitive functions (Cocco et al., 2018; Santos et al., 2018). Within this frame, more in-depth analyses of tDCS action are important to comprehending the molecular and neurophysiological mechanisms subtending tDCS effects on cognitive processes and DPMS dysfunction (Soldatelli et al., 2021). So, comprehension of its impact on the neuroplasticity processes, with the perspective to link them with clinical effectiveness, might help with better use of this technique.

Thus, we determine whether 20 sessions of a-tDCS on the left (l)-DLPFC and cathodal on the right (r)-DLPFC over 4 weeks self-applied at home would be superior to a sham-(s)-tDCS in improving the executive attention, divided attention, working memory, and cognitive flexibility assessed by the Trail Making Test (TMT-B-A) (primary outcome). Additionally, we evaluated its impacts on executive functioning (Controlled Oral Word Association Test; COWAT); the WM (Digits subtest from Wechsler Adult Intelligence Scale; WAIS-III); and quality of life (secondary outcomes). We investigated if the tDCS effects were related to the severity of the DPMS dysfunction at the start of the treatment and with neuroplasticity changes evaluated by the percent change in the BDNF from pre- to treatment end. We hypothesized that a-tDCS could improve cognitive performance more effectively than s-tDCS. Besides, we investigated whether these effects were correlated with the degree of pain processing pathway malfunction as measured by baseline DPMS deficit. We also investigated if the tDCS effects are mediated by changes in the neuroplasticity state, as indexed by serum BDNF.

This trial demonstrated that the current protocol of home-based a-tDCS compared to sham for 4 weeks of stimulation over the l-DLPFC improved the cognitive functions assessed by TMT-B-A (executive attention, divided attention, W.M., and cognitive flexibility). Additionally, they provide evidence of the a-tDCS benefits to improving secondary outcomes, including W.M., verbal semantic fluency, phonemic fluency, and the impact of F.M. symptoms on quality of life. The a-tDCS effects on the cognitive tests were positively correlated with the percent changes in averages from pre-treatment to treatment end of the BDNF. Also, the severity of dysfunction of DPMS at baseline predicted more remarkable a-tDCS effects in the cognitive impairment. Besides, we found that the reduction in the BDNF related to the a-tDCS is related to improving symptoms due to F.M. The study had a dropout rate of 10%, mainly due to restrictions on circulation in the streets instituted during the COVID-19 pandemic. Mild to moderate adverse events were more common in the active tDCS group, particularly skin tingling, burning, and itching, and the global adherence was 83.03%.

This trial has key methodological differences compared to previous studies on tDCS to improve the cognitive impairment in F.M. We used a home-based tDCS device that enabled a considerably higher number of sessions. Hence, until we can be known, this is the study that evaluated the highest number of sessions used to improve cognitive performance at home. This is particularly relevant since preliminary evidence points to the increased efficacy of tDCS with more extended periods of treatment (Brunoni et al., 2016; Castillo-Saavedra et al., 2016; Brietzke et al., 2020). Additionally, the need for daily visits to clinics or hospitals has always been a significant challenge for using tDCS in the clinical context (Charvet et al., 2020; Salehinejad et al., 2021). Thus, the home-based device opens a new window of opportunity, especially for subjects with physical or cognitive disabilities that hinder their access to the clinical center. So, these results corroborate other previous studies which found that the a-tDCS on the DLPFC might activate regions associated with pain processing, such as the anterior cingulate cortex (ACC) cortex, the primary somatosensory cortex (S.I., SII), insula, and thalamus (Apkarian et al., 2005; Leknes et al., 2008; Bushnell et al., 2013). The fact that pain, as well as attention, share the same cognitive network results in a hindrance to having an efficient cognitive system. Therefore, pain may impair voluntary attentional systems and associated executive functions (Eccleston and Crombez, 1999; Bell et al., 2018). A-tDCS can alter the electrical activity of specific brain regions, encourage cortical plasticity, and enhance functional connections in the area that is being treated, improving pain modulation and quality of life. So, the a-tDCS’s effect modulates neuronal membrane potential on cortical and subcortical neural networks involved in cognitive functions and pain processing. This effect corroborates the results of an earlier trial in a single session of tDCS with 2 mA, applied to the DLPFC cortex in F.M., which found improvements in the function of neural networks involved in spatial and executive attention, as well as a reduced perception of pain (Silva et al., 2017). Another trial also observed the benefit of eight tDCS sessions paired with cognitive training on working memory, verbal fluency, and immediate and delayed memory (Santos et al., 2018). Besides, the current findings are aligned with previous studies in patients with a depression diagnosis, which found that a-tDCS over the DLPFC reduced depressive symptoms and other symptoms linked to inappropriate emotional functioning (Brunoni et al., 2017). As well it reduced pain catastrophizing (Caumo et al., 2022) and improved cognitive functions (e.g., decision-making) (Dixon et al., 2017). According to the literature, this effect can be related to top-down control that up-regulates reactions to positive emotional stimuli (Grimm et al., 2008; Goldin et al., 2009). The ability of a-tDCS on the l-DLPFC to alleviate cognitive abnormalities, notably hypoactivity in the l-DLPFC and hyperactivity in the r-DLPFC, may be one of the potential mechanisms behind these processes and related to its impact on cognitive impairment. This hypothesis is supported in a study that assessed how inter-hemispheric connectivity conservation could have cognitive implications (Krupnik et al., 2021). So, the current result might contribute to a greater understanding of the tDCS effect on brain function.

According to the CPM-test, the severity of DPMS inhibitory dysfunction predicts a remarkable a-tDCS effect compared to s-tDCS on improving cognitive function. This finding suggests that the a-tDCS impact on the outcomes has been more evident in more severe diseases. These findings demonstrate that there is an interaction between the spine-bulbospinal loop and the neural network of cortical areas from an integrative approach. They support the notion that the DPMS and the brain networks involved in cognitive processing have similar neurobiological workings. According to the research, the DLPFC is, therefore, a crucial brain area for modulating the experience of pain. The benefits of using the l-DLPFC as a target area to modulate pain corroborate meta-analysis data that showed the a-tDCS on pain with a moderate E.S. (0.54) (Zortea et al., 2019). Besides, the DPLCF as a target area for improving cognitive performance finds support in data that links prefrontal cortex function with a decline in cognitive abilities (Wen et al., 2011; Wiseman et al., 2018), as well as with the impact of a-tDCS on the l-DLPFC in W.M (Santos et al., 2018). Thus, it is plausible that the cognitive impairment in chronic pain encompasses dysfunctions in neural networks in brain areas with a central role, either in pain (Staud and Spaeth, 2008; Bosma et al., 2016) or in cognition, such as the insula, ACC, and PFC (Nijs et al., 2021). Other studies showed the benefits of a-tDCS on l-DLPFC are supported by improvement in the W.M. and clinical and experimental pain either by repetitive transcranial magnetic stimulation (r-TMS) (Graff-Guerrero et al., 2005; Borckardt et al., 2007) or a-tDCS (Santos et al., 2018). This information reveals that the downstream regulating circuits, including the anterior insula, hypothalamus, periaqueductal gray substance, nucleus accumbens, and rostroventral medulla, are involved in the processes encompassing the effects of a-tDCS on the l-DLPFC (Wager et al., 2013).

The effect of repetitive sessions of a-tDCS has been attributed to the induction of use-dependent neuroplasticity, which is related to “synaptic learning” and long-term changes, which resemble glutamatergic synapses’ long-term potentiation (LTP) or long-term depression (LTD) (Nitsche and Paulus, 2000, 2001). The activity level of underlying neuronal populations at stimulation time is a potentially important mediator of the effect of tDCS on brain function. This is further corroborated by the fact that the impact of tDCS to improve cognitive performance is positively correlated with the neuroplasticity state, according to the percent change in serum BDNF from pre- to treatment end. This discovery aids in understanding how a-tDCS affects faulty neuroplasticity since it can alter mechanisms that include strengthening glutamatergic synapses while weakening GABAergic synapses (Coull et al., 2005). The relationship of serum BDNF to predict the a-tDCS was found in our previous studies with F.M. with a-tDCS applied to the DLPFC in work memory (Santos et al., 2018). In a study with a similar montage, the baseline BDNF predicted the tDCS effect on daily pain scores after sixty sessions of tDCS self-applied at home (Brietzke et al., 2020). Besides, in the postoperative recovery of the hallux valgus surgery, the liquor BDNF after two a-tDCS sessions was associated with lower pain scores and disability due to pain 7 days after surgery (Ribeiro et al., 2017). According to earlier studies, it has been found that there is higher serum BDNF in FM compared to other chronic pain and healthy subjects (Deitos et al., 2015; Stefani et al., 2019). Therefore, the decrease in serum BDNF in the a-tDCS group compared to the sham group and the improvement in cognitive function suggest that the intervention counter-regulated the FM-related dysfunctional neuroplasticity. Despite the relevance of this finding to indicate how much the effect of this therapy might help to improve maladaptive neuroplasticity, this result should interpret sparingly because it is an indirect measure of the neuroplasticity phenomenon.

Our findings should be viewed considering some limitations. First, although patients received comprehensive training in using the device, no remote monitoring of sessions was performed. Therefore, there should be caution in direct comparison with studies with supervised electrode placement and exposure supervision. Second, we included only females to remove the potential bias due to sex since it has been found that in women, the a-tDCS over DLPFC produces a higher current flow to the frontal regions (Russell et al., 2017) and better performance in cognitive tasks than in men (Martin et al., 2017). Third, our findings are consistent with past research that supported this method of self-application for prolonged tDCS use at home. We also see similar outcomes to studies in which the therapy was administered under close observation (Santos et al., 2018). Fourth, the tDCS system used in the current study provides an effective technical solution that enables medical engineers who were not involved in the patients’ care to program the tDCS system by the randomization sequence to ensure that all members of the research team and patients are blinded. Fifth, in this study, high adherence was observed by the records of devices in use, like those obtained in real-life environments. Sixth, despite the randomization processes permit to have balanced groups of a-tDCS and s-tDCS on cognitive performance may be confounded by other variables, such as psychiatric comorbidities (i.e., depression, anxiety, and sleep disturbance) (Austin et al., 2001; Airaksinen et al., 2005; Castaneda et al., 2008), or medication use, particularly opioids, which may lead to cognitive side effects that cannot be controlled entirely (Ersek et al., 2004). Seventh, there is no standard battery of neuropsychological tests for cognitive function assessment in chronic pain. So, the literature has recommended that the cognitive assessment in FM. should include tests to evaluate attention and W.M., complex psychomotor speed, and executive functioning (Kravitz and Katz, 2015). Eight, the allocation sequence was developed following a standard format described in the scientific literature. Table 1 reveals that most baseline variables are balanced across groups, indicating that randomization equilibrated the groups (a-tDCS and s-tTDC). Although the CSS baseline disparity between a-tDCS and s-tDCS may be attributable to random chance, it is not possible to rule out the effect of regression on the mean. That is, a higher score on the outcome initial might tend to be lower upon subsequent measurement. Nine, we decide by allocation 2:1 based on the rationale that fibromyalgia has important suffering. Based on the argument that fibromyalgia causes significant long-term discomfort, we allocate 2:1 since, if a lower number of participants in the sham group, we may treat more individuals actively, leading to increased adherence. Additionally, the higher sample size in the active group increases the ability to identify side effects (Dumville et al., 2006; Hey and Kimmelman, 2014). Finally, with an adherence rate of more than 85% to sessions in both a-tDCS and s-sham tDCS, we adopted a strict and reproducible technique to demonstrate the efficacy and viability of t-DCS at home. However, further studies must explore if neurophysiological measures, such as EEG records, might help to shed light on the specific modulated cognitive processes by the intervention. Another aspect is to allow a more focal target area of the stimulation using multichannel tDCS montages.

These findings revealed that daily treatment with a home-based tDCS device over l-DLPFC compared to sham stimulation over 4 weeks improved the cognitive impairment in F.M. The a-tDCS at home was well-tolerated, underlining its potential as an alternative treatment for cognitive dysfunction. Besides, the a-tDCS effect is related to the severity of DPMS dysfunction and changes in neuroplasticity state.

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

The Research Ethics Committee approved the protocol for this trial at the Hospital de Clínicas de Porto Alegre (HCPA), Brazil. The patients/participants provided their written informed consent to participate in this study.

WC, RA, PVS, and FF had substantial contributions to the conception and design of the work. CA, LR, PRS, DS, IL, FF, and WC drafted the work and revised it critically for important intellectual content. WC and FF agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work that were appropriately investigated and resolved. All authors agreed and approved the final version of this work.