The study was designed as a randomized, double-blinded placebo-controlled study. Forty-three subjects were recruited from the inpatients and outpatients of the Department of Rehabilitation at Gwangju Veteran Hospital between August 2021 and May 2022. We included the following subjects: (1) patients aged ≥19 years, (2) patients who had experienced cerebral infarction for the first time, (3) patients with unilateral cerebral infarction, (4) impaired balance (Berg Balance Scale [BBS] ≤45), (5) patients with functional ambulation categories ≥3, who could walk on their own, and (6) patients with a Mini-Mental State Examination (MMSE) score of ≥24 who could understand and respond to questionnaires. We excluded those (1) who could not undergo rTMS (history of seizures or epilepsy, intracranial metallic or magnetic material, pacemakers, and other implantable medical devices, taking tricyclic antidepressant (TCA) and neuroleptics, pregnant women), (2) with balance impairment due to reasons other than cerebral infarction (fractures, peripheral nerve damage, visual impairment, vestibular dysfunction or disorder), (3) those where the infarction was located in the cerebellum or PCA territory area, and (4) those with cognitive disability that impeded responding to the questionnaires. Among the 43 patients, 32 patients were included, excluding those who had cerebral infarction in the cerebellum or PCA territory region (n=3), who could not understand the questions due to cognitive impairment (n=7), and who could not undergo rTMS due to a pacemaker (n=1). Among these 32 patients, 16 patients were randomly assigned to the rTMS group while the other 16 patients were assigned to the control group (Fig. 1). The therapist who conducted rTMS and the personnel who performed outcome measurements were different, thereby making the protocol double blind.

This study was approved by the Institutional Review Board of Gwangju Veteran Hospital (No. 2021-17-2). The rights of all study participants were protected in accordance with the ethical principles of the Declaration of Helsinki and informed consent was obtained in advance.

Patients in both groups underwent a conventional rehabilitation program five times a week. Each day, patients performed passive and active mobilization exercises, gait training, balance training, and muscle strengthening for both upper and lower extremities for 30 minutes, followed by 30 minutes of training in grasp power, hand dexterity, and activities of daily living.

In the rTMS group, the coil of the machine used for treatment was placed 2 cm below and 2 cm lateral to the inion by targeting the cerebellar hemisphere contralateral to the site of cerebral infarction [13] (Fig. 2). In one session, 900 pulses of 1 Hz at 90% of the resting exercise threshold (RMT) were delivered for 15 minutes [14]. We determined the RMT by stimulating a motor hotspot with the lowest energy capable of generating at least 5 evoked potentials ≥0.05 mV within 10 stimuli [18]. Sessions were conducted five times per week for 2 weeks (total of 10 sessions). For rTMS treatment, Magpro R30 (Magventure, Farum, Denmark) and a figure-of-8 coil were used. It was performed by an experienced therapist. For the control group, a sham coil that did not induce a magnetic field was used, although the sound and scalp sensation were similar to those produced by the real rTMS coil [19].

The BBS score was used as the primary outcome measure. BBS is an objective measure of dynamic balance taken while performing 14 functional tasks. Items in the BBS are scored on a 4-point scale, with a maximum combined score of 56, where higher scores indicate better balance and functional independence with respect to the activities tested [20].

The 10-m walk test (10mWT), Timed Up and Go test (TUG), and Activity-specific Balance Confidence scale (ABC) scores were used as secondary outcome measures. The 10mWT is a gait assessment that determines gait speed. A stopwatch is used to measure the time taken to walk 10 m and gait speed is calculated by dividing the distance walked by the time taken (m/s). The higher the 10mWT value, the slower is the gait speed [21]. In the TUG test, the patient begins in a sitting position in a standard armchair (seat height of 46 cm). The patient is then instructed to rise from the chair, walk 3 m (usually with a walking aid) to colored tape placed on the floor, turn around, return to the chair, and sit back down again. The stopwatch is started when the patient's hips leave the seat and is stopped when the hips touch the seat again. A higher TUG value indicates a lower gait speed [22]. The ABC scale consists of 16 items, each representing a variety of indoor and outdoor activities that require different levels of balance skills. Respondents are asked to indicate their level of confidence in performing each activity, using a scale distribution from 0% to 100%, and the values are averaged. Higher scores indicate a higher confidence in performing the activity [23,24]. Assessments were performed by the patient’s attending therapist, who did not know if the patient was in the rTMS or control group. Depending on the timepoint, all assessment items were recorded at baseline (T0), after 10 sessions of rTMS (T1), and at 4 weeks after treatment completion (T2).

All statistical analyzes were performed using R software, version 4.0.1 (R Foundation for Statistical Computing, Vienna, Austria; http://www.R-project.org). Statistical significance was based on a p-value of <0.05. For continuous variables, normality of data distribution was tested using the Shapiro-Wilk test. If assumptions of normality were satisfied, repeated-measure analysis of variance (ANOVA) was used. If not, the Wilcoxon test was performed. In addition, two-sample t-test and Mann-Whitney U test were used to compare the rTMS group and the control group.