It is a triple-blind, sham-controlled crossover clinical trial following the protocol published by Uehara et al.16 This study follows the Consolidated Standards of Reporting Trials (CONSORT) statement. This trial was approved by the Research Ethics Committee of the Nove de Julho University (CAAE certificate: 26427819.1.0000.5511) and was registered in the Brazilian Clinical Trials Registry (ReBEC) (RBR-8zpnxz). All participants signed an informed consent document before participating in the study.

Participants were invited to participate in the study by advertisements through social networks and contacting recreational runner groups in the region where the study was conducted. Adult recreational runners who ran from 5 to 21 kilometers, three times a week, for at least six months were considered eligible for this study. Pregnant women, people who practiced other sports, who had recent musculoskeletal injuries, recurrent seizure history, skin lesions on the stimulation area, who used antihistamines, antidepressants, antiepileptic, pacemakers, or who had metal implants in the head were excluded from the study. Participants were instructed not to modify their training during the study.

Participants attended the research laboratory three times, with an interval of seven days between each session. In the first session, we collected the participants’ demographic data. In the second and third sessions, participants received the intervention protocol consisting of the anodal tDCS or the sham tDCS followed by the fatigue protocol, according to the randomization schedule (Fig. 1). Participants were instructed to avoid consuming caffeine, tea, alcohol, tobacco, and drugs and not perform physical activities the day before interventions.

Fig. 1.

Experimental study timeline. a-tDCS, active tDCS; s-tDCS, sham tDCS; MEP, motor evoked potential; RPE, rating of perceived exertion.

(0.19MB).

Participants were classified as responders or non-responders to the tDCS before the protocol. Motor-evoked potential (MEP) was measured immediately before and after anodal tDCS stimulation at 2 mA for five minutes. Then, those participants with initial and final MEP ratios greater than 1.0 were considered responders.17

The anodal tDCS was applied before the fatigue protocol, using the NeuroConn DC-STIMULATOR (Germany) for 20 min, 2 mA intensity, 30-s linear ramp up/down, using two non-metallic surface electrodes, 5 × 7 cm (cathode) and 5 × 5 cm (anode). The anode electrode was positioned over the dominant quadriceps muscle hotspot, previously established using TMS.16 The cathode was placed on the supraorbital contralateral region. For the sham tDCS, the same set up used for anodal tDCS was applied to the sham tDCS, except that the equipment was turned off after 20 s.

The fatigue protocol follows the methods by Schwendner et al.18 The participants were placed in an isokinetic dynamometer (Biodex System 2), having a 100° angle between the trunk and hip, and the lower dominant member at a 60° flexion (0° corresponding to full knee extension). The dynamometer axis was placed at the knee´s rotational axis.

The participants executed flexion and extension concentric contractions of the dominant limb knee, at 100° of maximum voluntary contraction (MVC), at a 60°/s speed , with a 60° range of motion (between 90° and 30° knee flexion). For each contraction, there was an opposing force provided by the dynamometer. Muscular fatigue was considered when the participant did three contractions at less than 50% of MVC (pre-established in isokinetic dynamometer).

Peripheral fatigue was assessed by collecting information on the following variables: peak torque, blood lactate level, and RPE. Central fatigue was determined by assessing the cortical excitability via MEP. The peak torque, MEP, and RPE were performed before and after the intervention protocol. Blood lactate was evaluated before, during, and after the fatigue protocol.

The sample size was calculated using the GPOWER3 software. The paired t-student test was used considering the difference in peak torque results between post and pre-tDCS intervention of five participants in the anodal tDCS (44.66±15.25 N.m) and sham tDCS (24.96±10.64 N.m). Based on the difference between groups, using a power of 90% and an alpha level set at 0.05, 16 participants were required. However, 30 participants were recruited to allow for possible dropouts.

Participants were randomized to receive the anodal tDCS or the sham tDCS in the first session. In the second session, participants received the other intervention. An investigator not involved in providing treatments or assessing participants was responsible for the randomization process. Randomization was performed using the website www.randomizer.org.

This study was triple-blind (assessors, therapists, and participants). Assessor blinding was possible because the assessors who measured the outcomes were blinded to the treatment received by each participant. The NeuroConn DC-STIMULATOR device has a configuration in which, through codes, the active or simulated mode is selected, so the researcher, who conducted the treatment, and the patient, did not know which mode was chosen. To validate the blinding of participants,22 participants answered at the end of each intervention the following question: “What type of treatment did you receive, the sham or the active tDCS?”.

Statistical analyses were performed using the SPSS (IBM; v.25.0). Only participants who completed both intervention sessions were considered for analysis. Data on peak torque, MEP, blood lactate level, and RPE were analyzed considering within-group change and between-group comparison. Data were presented as estimated marginal means and 95% confidence interval (CI). The statistical test used was Generalized Estimated Equations followed by Bonferroni post-hoc test for multiple comparison adjustments. For peak torque and MEP amplitude, two fixed factors were used: group [anodal tDCS, sham tDCS] x time [pretest, posttest]; for blood lactate two fixed factors were used: group [anodal tDCS, sham tDCS] x time [pre, during, post fatigue protocol]; for RPE the Wilcoxon test was used. Statistical significance was set at p < 0.05.

The reduction in peak torque after the fatigue protocol was predictable because physiological changes occur after fatiguing tasks, for example, changes in pH and cellular metabolites. These changes lead to a reduction in contractility and muscle strength.8,28 Our study hypothesis was that anodal tDCS applied to the quadriceps hotspot before the fatigue protocol would minimize fatigue (i.e., less reduction in peak torque for the tDCS group anodic); however, this hypothesis was not confirmed. However, Codella et al23 observed an increase in power and endurance of the lower limbs during running in healthy physically active men after tDCS, and Park et al24 observed an increase in exercise duration after tDCS was applied during running up to exhaustion. Other studies11,29,30 also demonstrated a reduction in fatigue assessed through time to exhaustion and motor failure tests; however, they were performed on healthy, physically active participants.

These results may have occurred because tDCS can modulate cortical excitability, increasing the neuronal drive of the locomotor muscles located in M1.29 However, for this effect, more than a single session of tDCS may be necessary, as reported by some studies31–34 which indicates that a single session of tDCS in M1 failed to improve maximal anaerobic exercise. Giboin and Gruber25 also did not observe the effects of tDCS on knee extensor strength applied during and before the fatiguing task. This result may occur because under conditions of maximum strength, the muscles are already working at maximum, so all motor units are already recruited; therefore, ceiling effects do not allow tDCS to show additional effects.12,13

Another factor to be considered is that we used a 2 mA intensity, as in other studies with tDCS and fatigue.11,25,29 However, Workman et al35 observed better effects on peak torque after the fatigue protocol in physically active adults when tDCS was applied at 4 mA compared to 2 mA. Despite this, few studies investigate the stimulation's effects with intensities greater than 2 mA. Therefore, further studies with higher intensities should confirm whether stimulation intensity could interfere with tDCS’ effects on fatigue.

There was a significant increase in lactate 2 and 3 compared to lactate 1 levels in both groups, which was expected, because lactic acid tends to increase proportionally to exercise intensity, being considered a fatigue marker36,37 However, as there was no difference between the groups, the anodal tDCS did not interfere with the lactate level. The same was observed in elite swimmers who received tDCS.38

The RPE results of the groups in our study increased, without significant differences between them. Similarly, Baldari et al39 did not observe the tDCS’ effects on the RPE of recreational runners during the incremental ramp test. However, Williams et al30 observed improvements in RPE in healthy participants who received anodal tDCS during elbow flexors submaximal sustained contraction. The difference in results may be because previous studies evaluated smaller muscle groups; therefore, the exertion feeling would seem more easily recoverable.

The MEP results show a reduction in both groups, which is expected after fatiguing tasks because there is a decrease in the motor neurons firing due to the excitability depletion or in the excitatory synapse response.5

However, we hypothesized that anodal tDCS applied over the quadriceps muscle hotspot could reduce fatigue compared to the sham tDCS, but we did not find these results. In contrast, Angius et al33 observed in active men an increase in MEP amplitude after anodal tDCS in the knee extensors isometric exhaustion test, and Kristiansen et al40 observed an increase in cortical excitability after anodal tDCS in physically active individuals, but without change on fatigue time on bike and RPE.

We argue that our findings regarding MEP are due to individual variability, as our participants were not homogeneous concerning functional capacity and kilometers covered in the race. Furthermore, of the 30 participants, only seven were tDCS responders. Therefore, repeating this protocol with more participants responsive to the stimulation technique may produce different results. Another hypothesis is that in a longitudinal study, we may find different results, as observed by Lopez-Alonso et al,17 who evaluated intra- and inter-subject variability in healthy individuals and observed a 44% increase in responders to tDCS after the second session compared to the first session. In addition, the literature shows us that the tDCS’ effects are summative and stimulate neuroplasticity.41 This effect was demonstrated by Ammann et al42 who reported an increase in the healthy individuals’ MEP amplitude from the first to the third tDCS session; moreover, longitudinal evidence of performance improvement through the M1 tDCS was provided in adolescents’ professional rowing athletes.43

However, our view is that further research is still needed to investigate whether tDCS is for everyone. Esteves et al44 report that the training level can affect neuronal activity and brain structure, being able to produce different responses to stimulation. It has been shown that tDCS has more effect when individuals have a lower baseline activity or experience, i.e., when changes in synaptic connectivity are required; therefore, when there is the possibility of neuron functional improvements. Despite being recreational runners, almost all participants in the current study practiced exercises more than five times a week, which may have altered the motor threshold and cortical excitability, making it laborious to obtain a response. Performance improvement becomes increasingly arduous to achieve as the performance level rises due to the so-called “ceiling effects”.45

A recent review46 that included 19 studies looked at the tDCS acute effects on changes in athletes' motor performance compared to simulation and showed that anodal tDCS led to better performance in athletes on sport-specific motor tasks, but none of the athletes in the review were recreational runners. Of the 19 studies included, 10 investigated resistance effects; five found an increase in specific endurance performance, while five showed nothing.

Therefore, we believe that, for clinical practice, more studies are necessary to evaluate the participants' neurophysiological functions who received more than one intervention session, considering the variability that can interfere with the results, such as age, sex, skull, and brain shape.47 Thus, individualized stimulation protocols are necessary to understand how tDCS works for athletes. As with other therapies and medications, for example.

Finally, a last hypothesis for not finding a tDCS effect would be the rise in cortisol and glycogen levels. Although we did not perform this measurement, Mellow et al48 state that after high-intensity exercises, as the fatigue protocol, there is an increase in the cortisol and glycogen levels that can block the non-invasive stimulation effects on neuroplasticity. Thus, these measures may be relevant for future studies involving tDCS and fatigue.

As limiting factors for this study, we can highlight the participants’ heterogeneity concerning the distance they run, as we had eight 5 km runners, eleven 10 km runners, and eleven 21 km runners. Furthermore, they presented different life and training history. López-Alonso et al17 reported this same difficulty, noting a large inter-individual variability in the response to tDCS, explaining the difference in results.

Of the 30 participants, only seven were tDCS responders. We believe that this study reinforces the individual´s evaluation importance and that future studies evaluate the longitudinal tDCS effects within the specific functions of each sport.

However, it is necessary to note that brain stimulation may be considered as a “neuro-doping” agent, as it can improve physical and mental performance in sports. Therefore, with the growing field of brain stimulation for performance, this question may be one relevant point for debate.