A parallel randomized clinical trial was conducted with a double-blind placebo control. Sixty healthy subjects volunteered to participate. All participants provided written informed consent, which had been approved by the relevant local ethics committee (Ref: 441; 11/11/2019). The study was formerly registered at ClinicalTrials.gov (NCT04346719). Epidat 3.1 software was employed for the simple balanced randomization of the sample. Subjects were randomly assigned to the three intervention groups (10 kHz HFAC, 20 kHz HFAC, and sham stimulation) by an independent investigator. Both participants and assessors were blinded to the group assignment, which was kept in a closed envelope throughout the intervention so that only the researcher who delivered the intervention was aware of group allocations. The intervention lasted 20 minutes. Variables were measured at four time points: i) pre-intervention (0 min), ii) during the intervention,15 minutes following the start of the stimulation (15 min), iii) immediately post-intervention (20 min), and iv) 15 minutes after the finalization of the intervention (35 min) . However, both the antidromic sensitive action potential (SNAP) and maximal flexion finger strength (MFFS) were not recorded during the intervention because of interferences with the application of electric currents and the discomfort needles cause during muscle contractions, respectively.
Healthy volunteers were recruited, 18 to 40 years of age, with no pathologies of the nervous system, and with no allergy to nickel or intolerance to the percutaneous application of electric currents. The criteria for exclusion were: surgical procedure or osteosynthesis material in the upper limb where the electric stimulation is to be applied, epilepsy, fear of needles, infectious disease, neuro-muscle disease, heart failure, diabetes, cancer, pacemaker or other implanted electric devices, pregnancy, tattoo or skin condition in the area not allowing for the delivery of the intervention, and use of substances or medication (e.g. anticlotting, thrombolytic, analgesic, corticoid, antidepressant, antiepileptic) during the trial and in the seven days before their participation.
The duration of both active HFAC and sham stimulation was 20 minutes. The interventions were delivered to the non-dominant arm with the participants in the supine position. Antiseptic and skin disinfection treatment with 2% chlorhexidine in an alcohol base was applied on the intervention area. A Samsung HS50 (Samsung healthcare; Seoul, South Korea) ultrasound device was employed for the percutaneous guided application of currents with a lineal probe of 12mHz. A short-axis approach to the median nerve was performed on the anterior aspect of the middle third of the forearm placing two 0.30 mm x 40 mm acupuncture needles (Agupunt®; Barcelona, Spain) close to the epineurium of the median nerve (1 mm), one needle on each side of the nerve. The average depth of the needle introduced into the tissue was 3 cm (Figure 1).
A Myomed 932 (Enraf-Nonius; Delft, The Netherlands) device connected to the needles by a clamp applied the current in the three interventions. All interventions were delivered in a university laboratory facility under reduced noise and temperature (21ºC–25ºC) conditions.
10 kHz and 20 kHz current stimulation
High frequency alternating currents (HFAC) with sinusoidal waveform were applied at a frequency of either 10 kHz or 20 kHz in the active intervention groups. The intensity progressively increased until producing a feeling of “strong but comfortable” tingling, just under the motor threshold . Subsequently, the intensity was increased until a minimal visible contraction was observed and then slightly lowered below the motor threshold. Due to the accommodation to the stimulus, the intensity was adjusted every two minutes to rise it if the participant’s perception of the current decreased .
Sham stimulation was applied with the same device and needle placement. The same parameters were used as in the 10 kHz intervention except for the current intensity, which was initially adjusted up to the sensitive threshold and, once the participant perceived a tingling sensation for a few seconds, the intensity was gradually lowered to and maintained at 0 mA for the entire session.
D. Outcome Measurements
Maximum isometric finger flexion strength and myotonometry
The main outcome measurements related to motor activity were the MFFS of the index finger and mechanical characteristics of the opponens pollicis muscle as assessed through myotonometry. MFFS of the index finger was evaluated with the participant in the supine position and their hand in pronation pressing on a MicroFet 2TM digital hand dynamometer (Hoggan Scientific, LLC; Utah, USA), a device with proven intra- and inter-assessor reliability . The MFFS was calculated as the mean of three measurements in kgs that were taken with a contraction time of 3 seconds and a rest of 5 seconds between measurements .
A myotonometer (Myoton AS; Tallinn, Estonia) was used to evaluate the mechanical properties of the muscle. Ten mechanical stimuli of 0.4 N force and 0.15 ms duration were applied to the opponens pollicis muscle of the limb where the intervention was performed with one-second intervals between stimuli. If the variation coefficient exceeded 3%, the measurement was repeated. Stiffness (N/m), frequency (Hz), and logarithmic decrement (expressed in arbitrary units) were the collected variables. Stiffness measures the force that results in tissue shape changes. The frequency of damped oscillations serves to measure the resistance of the tissue to mechanical stress and is considered to be an indirect measure of muscle tone. The decrement serves to characterize tissue elasticity by measuring the loss of mechanical energy as the amplitude of the oscillations decreases [26-27].
Mechanical detection threshold and pressure pain threshold
Somatic sensitivity was evaluated by means of the mechanical detection threshold (MDT) and pressure pain threshold (PPT). The MDT was measured via modified Von Frey filaments (OptiHair2, MARSTOCKnervtest; Marburg, Germany) on the palmar aspect of the hand in an area of 1cm2 proximal to the head of the second metacarpal and on the thenar eminence. Filaments with a diameter of 0.4 mm delivered forces of 0.25, 0.5, 1, 2, 4, 8, 16, 32, 64, 128, 256, and 512 mN . Seven stimuli were applied and the threshold was determined when at least four were perceived with a filament .
The PPT was recorded on the palmar aspect of the trapeziometacarpal joint via a digital algometer with an increment scale of 0.1 N (Wagner Instruments, model FDIX; Greenwich, USA) and a circular applicator of 1 cm in diameter. The pressure was increased at an approximate rate of 5 N/s . Three measurements were taken with an interval of 10 seconds between them  and the PPT (N) was obtained from the average of the three measurements [32-34].
Antidromic sensory nerve action potential (SNAP)
The SNAP of the median nerve was recorded for assessing the effect on peripheral nerve conduction . Nerve stimulation was performed on the inner side of the arm using a transcutaneous bipolar electrode, with a fixed distance between electrodes of 1 cm and placing the cathode 40 cm from the recording electrode. Two ring electrodes on the index finger were employed to record the potential, with the ground electrode placed on the radial side of the wrist joint [36-37]. A constant-current stimulator (Digitimer LTD, model DS7A; Letchworth Garden, United Kingdom), an analogic/digital data acquisition card (Cambridge Electronic Devices; Cambridge, United Kingdom), and an amplifier (ETH-256 iWorxs; Dover, USA) with a 3Hz high-pass filter and a 2000Hz low-pass filter and an amplification of 1 were employed for the stimulation and recording. Supramaximal stimuli were applied with a pulse width of 1000 µs and a frequency of 1Hz. The latency and amplitude of the SNAP were calculated as the mean value of ten measurements. At baseline, two SNAPs were recorded with a 2-minute interval to analyze the power stability, and the average was used as the basal SNAP value.
Temperature of forearm and arm
A temperature monitor (model DRT4, Moor Instruments brand; Devon, United Kingdom) was employed to record temperature. One recording sensor was applied distal to the procedure, on the palmar side of the head of the first metacarpal, and another sensor proximal to the procedure, on the anterior side of the forearm . The room ambient temperature was also recorded.
Adverse effects and subjective perception
For the evaluation of adverse effects and the subjective perception of the participants, a standardized questionnaire was designed and completed at the end of the intervention. The questionnaire included nine items with "Yes/No” response options to evaluate pain, swelling, heat, redness, coldness, numbness, loss of strength, heaviness, and tingling in the hand and the intervention area. The unpleasantness and pain feeling perceived during the intervention were also assessed using a numerical scale from 0 to 10, where 0 corresponded to "not at all" and 10 to "the maximum possible". Additionally, participants were asked to report whether they perceived any of the above-mentioned effects or sensations in the intervention area in the days following the intervention.
Assessment of blinding success
The blinding success of the participants and the evaluator was assessed after the intervention ended . For this purpose, they were asked "What type of treatment do you believe you or the participant have received?", with five response options: (1) "I strongly believe that I have received an experimental treatment"; (2) "I somewhat believe that I have received an experimental treatment"; (3) "I strongly believe that I have received a placebo"; (4) "I somewhat believe that I have received a placebo"; (5) "Do not know, no answer".
E. Statistical Analysis
The sample size was calculated based on a previous pilot test carried out on seven healthy volunteers . For an expected between-group mean difference (MD) in the PPT of 10.3 N/cm2, with a standard deviation (SD) of 11.3 N/cm2 in the experimental group and SD 9.9 N/cm2 in the control group, and considering a type I error (α) of 0.05 and a power of 80%, the sample size was estimated to be 17 subjects per group (n=17). To compensate for possible dropouts, a supplementary 17% was added to the sample finally yielding a total of n=20 participants per group.
For the comparison of basal characteristics between groups, a descriptive analysis and inferential statistics for basal demographic variables were performed for independent groups (parametric or non-parametric depending on the variable). A two-factor (intervention-time) repeated-measures analysis of variance (ANOVA) with a Bonferroni post-hoc correction was conducted for the following outcome variables: MFFS, myotonometry, PPT, temperature, and SNAP. For those variables violating sphericity, the Greenhouse-Geisser correction was employed. Additionally, changes in the above-mentioned variables over time were calculated and an intergroup comparison was performed via a one-factor (intervention) analysis of variance (ANOVA) with a Bonferroni post-hoc correction. The Friedman test was employed for assessing the MDT with a post-hoc analysis via the Tukey’s test for intragroup comparison. The Kruskal-Wallis test was used for the comparison of MDTs between interventions.
The Chi-squared test was used for the analysis of adverse effects. Unpleasantness and pain during the intervention were evaluated by means of a one-factor (intervention) ANOVA with a Bonferroni post-hoc correction. All outcome variables were normalized in percentages with respect to basal values prior to the analyses and statistical significance was set at p<0.05. The IBM SPSS Statistic 24.0 software for Mac was used for all statistical analyses.