A threshold separates effective from ineffective dosing
The study's central comparison is not simply whether 1 Hz rTMS changes brain activity, but at which intensity it does so. Twenty healthy right-handed adults received four sessions in a randomized, single-blind crossover design: sham stimulation and 1 Hz rTMS over the left motor cortex at 80%, 100%, and 120% of each person's resting motor threshold (RMT), the lowest intensity that produced motor-evoked potentials of at least 50 µV in the right thumb muscle in at least 5 of 10 trials. Each active session delivered 1200 pulses over 20 minutes.
Compared with sham, stimulation at and above threshold produced robust transient responses: larger TMS-evoked potentials, higher beta-band power, higher beta-band lateralization, and recruited bilateral cortical activity. The 80% condition changed some evoked potentials but did not significantly change transient oscillatory power or lateralization, and it did not produce a significant aftereffect in lateralization. The paper therefore describes an intensity-dependent threshold under these experimental conditions rather than a smooth dose-response across the whole 80–120% range.
Transient lateralization and aftereffects are distinct measures
Two measurements carry the argument and should not be conflated. The transient lateralization index is computed from EEG recorded during the 20-minute stimulation period, capturing how beta-band oscillatory power is distributed between hemispheres while rTMS is being delivered. The aftereffect index is computed from resting-state EEG recorded for 10 minutes immediately before and immediately after stimulation, capturing the persistent shift that remains once stimulation stops.
The direction of change differs between the two windows. During stimulation at 100% and 120% RMT, beta-band lateralization increased, indicating stronger unilateral engagement. After stimulation, hemispheric lateralization decreased relative to the pre-stimulation baseline. The paper's interpretation is that the immediate response and the later rebalancing are linked, not that they are the same phenomenon measured twice.
Beta-band lateralization statistically mediates the aftereffect
The mediation analysis is the paper's main mechanistic claim. In a model comparing sham with the 100% and 120% conditions, stimulation intensity was associated with the aftereffect, and transient beta-band lateralization carried a significant indirect path, with an a path of 1.43 and a c path of 1.95, both marked as p < 0.05. A second model comparing sham, 100%, and 120% reported an a path of 0.60 and a c path of 1.16, again p < 0.05.
In plain terms, the stronger the unilateral beta-band engagement during stimulation, the larger the subsequent reduction in hemispheric lateralization. The authors describe this as the brain's immediate response being functionally linked to its therapeutic aftereffect rather than merely incidental. That is a statistical mediation result from an EEG association, so it supports a pathway hypothesis; it does not by itself prove that changing transient lateralization causes the aftereffect.
Evoked potentials and oscillations capture different responses
The time-domain and time-frequency analyses did not behave identically. Source maps of averaged TMS-evoked potentials showed focal, time-locked activation predominantly in the stimulated left motor cortex at 100% and 120% RMT. Beta-band oscillatory power maps, by contrast, showed bilateral activation patterns that scaled with stimulation intensity.
The authors read this divergence as evidence that the two measures capture partially different aspects of cortical reactivity, with oscillatory measures more sensitive to broader bilateral network engagement under these analytical conditions. This matters for anyone comparing rTMS studies: a result based on evoked potential amplitude and a result based on beta-band power are not interchangeable summaries of the same effect. The paper also notes that, unlike single-pulse TMS work reporting a linear intensity-to-power relationship, its repetitive protocol produced significant neuromodulation only at intensities at or above threshold.
A small stroke cohort links lower beta lateralization to better function
An exploratory cohort of 10 stroke patients received 10 sessions of 1 Hz rTMS together with conventional rehabilitation. Motor function improved over the rehabilitation period, and pathological interhemispheric asymmetry decreased. Across frequency bands, beta-band lateralization showed a significant negative correlation with motor scores, meaning lower lateralization values were associated with higher function; alpha- and theta-band relationships pointed the same way but were not significant.
This cohort is uncontrolled and has no sham condition, so it cannot establish that rTMS caused the improvement or that beta-band lateralization is a treatment biomarker. Its role in the paper is to motivate the healthy-participant finding: beta-band interhemispheric balance is proposed as a candidate neurophysiological correlate of motor recovery, not as a validated clinical predictor.
The evidence supports a marker hypothesis, not a dosing rule
Several design features bound how far the results travel. The healthy sample was 20 adults, enough for the prespecified power analysis but modest, with an unequal sex distribution that precluded testing sex-related effects. RMT was measured once before the first session and not reassessed, so day-to-day variability in corticospinal excitability could have reduced dosing precision. EEG was limited to 34 electrodes because the neuronavigation headband interfered with outer-ring electrodes, and impedances were kept below 10 kΩ rather than the stricter 5 kΩ some TMS-EEG recommendations prefer, so residual artifacts cannot be fully excluded.
The sham condition tilted the coil 90° at 100% RMT to mimic sound, sensation, and vibration without effective cortical stimulation, and participants wore earplugs. Even so, TMS-related sensations and discomfort were not formally quantified, leaving open some contribution from nonspecific sensory or affective responses. These constraints qualify the mechanistic reading: the mediation result identifies transient beta-band lateralization as a candidate marker for optimizing individualized dosing, but the study does not establish a clinical threshold for choosing an intensity in an individual patient.
The practical boundary sits at motor threshold
For readers weighing what to take from this paper, the most decision-relevant split is between subthreshold and at-or-above-threshold stimulation. At 80% RMT, 1 Hz rTMS over the left motor cortex did not significantly shift transient oscillatory lateralization or the subsequent aftereffect, while 100% and 120% did. The paper situates this against prior reports that subthreshold low-frequency rTMS can produce effects under other protocols, and concludes that subthreshold effects are protocol-dependent and may vary with the outcome measure used.
The priority the evidence supports is therefore measurement as much as intensity: if transient beta-band lateralization during stimulation tracks the later rebalancing, monitoring that signal could help distinguish a session that is engaging the intended network from one that is not. That is a hypothesis generated by a 20-person healthy cohort plus a 10-patient exploratory cohort, and it is the specific claim worth testing next, rather than a general endorsement of higher-intensity stimulation.
Sources used in this answer
Transient hemispheric lateralization is associated with intensity-dependent aftereffects of low-frequency rTMS
In 20 healthy adults, 1 Hz rTMS at 100% and 120% of resting motor threshold, but not 80%, produced transient beta-band lateralization that statistically mediated a later decrease in hemispheric lateralization; in an exploratory 10-patient stroke cohort, lower beta-band lateralization correlated with better motor function.