In the present study, we included a total number of 16 healthy subjects between 18 and 35 years of age (5 females). Subjects gave written informed consent to participate in the experiment according to the declaration of Helsinki and the ethics committee of the University of Leipzig approved the study. Prior to participation, all subjects underwent a comprehensive neurological examination. They were not taking any medication. Subjects that did not meet the protocol criteria and/or had contraindications for the study procedures were excluded from participation. According to the Oldfield questionnaire for the assessment of handedness , all subjects were right-handed. In a cross-over design, subjects were randomly allocated on the first day of the experiment into one of the following intervention groups: (A) conventional iTMS with an IPI of 1.5 ms (iTMS1.5 ms) and (B) iTMS with an individually adjusted IPI (iTMSadj). Both interventions were applied on separate days (reversed order, one week apart) in order to control for potential carry-over effects. Since the aim of the present study was to compare the effects of iTMS1.5 ms with iTMSadj, a total number of three subjects were excluded from data analysis since they showed an individual I-wave periodicity of 1.5 ms.
During the experiment, subjects were seated in an armchair with both arms relaxed and were instructed to keep their eyes open. Surface electromyogram (EMG) was recorded from the first dorsal interosseous (FDI) muscle of the right hand using surface Ag/AgCl electrodes in a bipolar montage. The signal was amplified using an EMG device (D360 8-channel amplifier, Digitimer Ltd, Welwyn Garden City, Hertfordshire, UK) with band pass filtering between 50 and 2000 Hz. The signal was digitized at a frequency of 5000 Hz (CED Power 1401, Cambridge Electronic Design, Cambridge, UK) and fed off-line to a data acquisition system (Signal Version 4.02 for Windows, Cambridge Electronic Design, Cambridge, UK) for further analysis. The absence of voluntary contraction during TMS was monitored online by visual inspection of the EMG signal and off-line by inspection of each individual trace. Trials with background EMG were excluded from the analysis.
TMS stimuli were delivered using a Magstim 200 (Magstim Co., Whitland, South West Wales, UK) through a figure-of-eight 70 mm coil. Initially, the position of the coil was identified over the motor cortex with the handle of the coil pointing posterolaterally with a 45° angle to the sagittal plane to elicit the largest and most consistent MEP amplitude in the right FDI hand muscle. This position was marked (= motor hotspot)  using a frameless stereotaxic neuronavigation system (Brainsight™, Montreal, Canada) and monitored online throughout the experiment to exclude any movement of the coil during the stimulation period. The motor hotspot of the FDI muscle representation was identified as the scalp position at which single TMS pulses at slightly suprathreshold intensity induced the most consistent MEP amplitudes in the relaxed muscle.
Resting motor threshold (RMT) over the left M1 was defined as the lowest intensity capable of evoking 5 out of 10 MEPs with amplitudes of at least 50 μV in the relaxed contralateral FDI muscle . TMS pulses were delivered at 0.2 Hz to the left M1 motor hotspot, a rate that does not affect cortical excitability at rest .
iTMS1.5 ms was applied over the left M1 as previously described . In brief, for iTMS1.5 ms paired-pulse TMS pulses of equal strength were delivered at an IPI of 1.5 ms and a repetition rate of 0.2 Hz for 10 minutes, resulting in a total number of 120 paired-pulse TMS stimuli. Stimulus intensity was initially adjusted to generate a MEP of approximately 0.5 - 1.0 mV resulting from paired-pulse TMS and was kept constant throughout the intervention. The reason for choosing an iTMS protocol (10 minutes) shorter than the "standard" 30 min iTMS intervention was to reduce the effective time of the study protocol.
For iTMSadj, we first measured the individual I-wave peak latency as indexed by the first I-wave peak (I1-wave peak latency) non-invasively by means of paired-pulse TMS as previously described [6, 7]. Stimulus intensity was kept constant for each TMS pulse, and adjusted to generate a MEP of approximately 1 mV when delivered alone, and no more than 4 mV when delivered as a 1.5 ms TMS pulse pair. A total number of 7 IPIs ranging between 1.2 and 1.8 ms in 0.1 ms steps (1.2, 1.3, 1.4, 1.5, 1.6, 1.7 and 1.8 ms) were recorded and compared to a single pulse TMS condition where test MEPs of approximately 1 mV amplitude were elicited. For each IPI and test condition a total number of 10 TMS pulses were recorded. The I1-wave peak latency for each subject was determined as the IPI in which the largest MEP facilitation out of all IPIs could be identified. Subsequently, iTMS was applied with the respective IPI (iTMSadj) while keeping the remaining parameters identical to the iTMS1.5 ms condition. Therefore, the only difference between iTMS1.5 ms and iTMSadj was the IPI between both interventions.
Data were analyzed using PASW for Windows version 18. In order to identify ppMEP facilitation in both iTMS interventions separately, we first performed a repeated measures ANOVA (ANOVARM) with factor TIME (1-10 min). Mean amplitudes of ppMEPs obtained in each minute (total of 10 ppMEPs per minute) of both interventions were calculated for each subject and expressed as a percentage of the mean data for the first minute . In order to compare the effect of iTMS1.5 ms and iTMSadj on ppMEP facilitation we performed an ANOVARM, if necessary with Greenhouse-Geisser sphericity correction, with factor TIME (1-10 min) and GROUP (iTMS1.5 ms and iTMSadj). Correlation analysis was performed using a linear Pearson correlation coefficient. Group mean data is presented as mean ± standard error of the mean (s.e.m.).