Low frequency rtms effects on sensorimotor synchronization

Transcription

1 Exp Brain Res (2005) 167: DOI /s RESEARCH ARTICLE Michail Doumas Æ Peter Praamstra Æ Alan M. Wing Low frequency rtms effects on sensorimotor synchronization Received: 12 January 2005 / Accepted: 28 March 2005 / Published online: 3 August 2005 Ó Springer-Verlag 2005 Abstract Previous studies using low frequency (1 Hz) rtms over the motor and premotor cortex have examined repetitive movements, but focused either on motor aspects of performance such as movement speed, or on variability of the produced intervals. A novel question is whether TMS affects the synchronization of repetitive movements with an external cue (sensorimotor synchronization). In the present study participants synchronized finger taps with the tones of an auditory metronome. The aim of the study was to examine whether motor and premotor cortical inhibition induced by rtms affects timing aspects of synchronization performance such as the coupling between the tap and the tone and error correction after a metronome perturbation. Metronome sequences included perturbations corresponding to a change in the duration of a single interval (phase shifts) that were either small and below the threshold for conscious perception (10 ms) or large and perceivable (50 ms). Both premotor and motor cortex stimulation induced inhibition, as reflected in a lengthening of the silent period. Neither motor nor premotor cortex rtms altered error correction after a phase shift. However, motor cortex stimulation made participants tap closer to the tone, yielding a decrease in tap-tone asynchrony. This provides the first neurophysiological demonstration of a dissociation between error correction and tap-tone asynchrony in sensorimotor synchronization. We discuss the results in terms of current theories of timing and error correction. Keywords Timing Æ Sensorimotor synchronization Æ Low frequency rtms Æ Motor cortex Æ Dorsal premotor cortex M. Doumas (&) Æ P. Praamstra Æ A. M. Wing Behavioural Brain Sciences Center, School of Psychology, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, UK mxd181@bham.ac.uk Introduction Repetitive transcranial magnetic stimulation (rtms) at low frequencies ( 1 Hz) decreases cortical excitability, an effect that can outlast the duration of the stimulation (for a review see Siebner and Rothwell 2003). Inhibitory effects of rtms have been studied mainly after stimulation over the primary motor cortex. Excitability changes of the primary motor cortex can also be induced by stimulation of areas with strong connections to the motor cortex such as the premotor cortex or the opposite hemisphere motor cortex (Gerschlager et al. 2001; Mu nchau et al. 2002; Rizzo et al. 2004; Kobayashi et al. 2004). Effects on cortical excitability are inferred from changes in the threshold for eliciting motor evoked potentials (MEPs) (Pascual-Leone et al. 1998; Maeda et al. 2000), changes in the duration of the cortical silent period (SP) (Ridding et al. 1995; Khedr et al. 2004), or from reduced or enhanced effects of conditioning as measured by paired pulse stimulation (Ridding et al. 1995; Modugno et al. 2003). Changes in excitability are not necessarily reflected in behavioural measures of motor performance. For instance, after 1 Hz rtms, basic movement parameters including peak force and acceleration (Muellbacher et al. 2000) and speed of finger tapping (Chen et al. 1997; Lee et al. 2003; Rossi et al. 2000; Sommer et al. 2002) remained unaffected, although one study found slowing of fastest finger movements (Ja ncke et al. 2004). In the present study we investigated effects of 1 Hz rtms on timing aspects of movement performance that have not been addressed in these previous studies. Of the studies that used repetitive tapping tasks to examine motor performance after rtms, two investigated paced tapping (Theoret et al. 2001; Ja ncke et al. 2004). The evaluation of motor performance in these studies was confined to interresponse interval measures, including their variability, but did not address whether rtms affected participants ability to synchronize their tapping to the pacing signal. In a sensorimotor

2 239 synchronization task, finger taps are synchronized with pacing stimuli provided by a metronome. A commonly observed phenomenon in this task is the tendency of the taps to precede the tones by up to 100 ms (negative asynchrony). Behavioural studies of paced tapping with different effectors have shown that foot tapping exhibits a stronger anticipation tendency than finger tapping (Paillard 1948; Fraisse 1980). This result suggests that the negative asynchrony is the result of a longer delay in sensory registration of the foot tap than the finger tap and that both feedback delays are longer than the delay in processing auditory information from the tone (Aschersleben and Prinz 1995; Aschersleben et al. 2002). However, other studies have shown that asynchrony decreases with interval and disappears at fast rates (Mates et al. 1994; Repp 2003), contrary to the feedback delay hypothesis. Another phenomenon observed in sensorimotor synchronization is error correction, a process necessary to monitor and correct deviations from temporal regularity. This process has been examined and modelled using isochronous paced tapping (Semjen et al. 2000; Vorberg and Wing 1996). According to these studies, synchronization involves a linear error correction process. Deviation from synchrony for one response is estimated, and a proportion of this deviation is used to correct the timing of the next response. An experimental approach to examine error correction in sensorimotor synchronization is to impose a local perturbation in an otherwise regular sequence of metronome stimuli (Mates 1994; Repp 2000, 2001a, b). Shortening or lengthening of a single interstimulus interval introduces a phase shift in the metronome sequence with a concomitant change in the temporal relation of taps and tones. Notably, the induced error is swiftly corrected, with a return to the baseline tap-tone asynchrony within a few taps. The correction of a stimulus phase shift does not depend on participants being able to perceive the shift and occurs for small shifts of 10 ms as well as for larger clearly perceivable shifts. Adequate correction of phase shifts that are below the threshold for conscious detection suggests that correction is achieved by means of an automatic lower-level phase correction process, possibly realized in structures close to the motor output such as the sensorimotor cortex. Involvement of the sensorimotor cortex in synchronization is supported by neuroimaging findings (e.g. Rao et al. 1997; Ja ncke et al. 2000). When the phase shift is clearly detectable, correction may also involve higher-level central timing mechanisms (Repp 2001a; Stephan et al. 2002; Praamstra et al. 2003; Repp and Keller 2004). To summarize, where previous rtms investigations of motor and premotor cortex have evaluated effects on behaviour mainly in terms of voluntary movement generation and timing variability, the present study investigated effects on sensorimotor synchronization. Evidence from previous research suggests that sensorimotor synchronization and error correction, especially correction of phase shifts that are not consciously detectable, are automatic processes possibly involving the contralateral sensorimotor cortex. We hypothesized that inhibition of the motor cortex by means of motor or premotor cortex rtms would affect synchronization and error correction, particularly the automatic correction of small perturbations, whereas with large perceivable perturbations error correction would be less affected. Participants and methods Participants Nine right-handed volunteers took part in this study, seven male and one female, (mean age=30, SD= 6.5 years). Handedness was scored on the basis of selfreport. They all participated in the task in which the motor cortex was stimulated. Six of them also took part in the premotor cortex stimulation task. All participants gave informed consent and the investigation was approved by the department s ethical review board. In one participant, the coil had to be replaced more than two times during premotor cortex stimulation, causing widely divergent effects of motor and premotor cortex stimulation on the silent period. Data from this participant were excluded from further analyses. EMG recordings EMG was recorded from the first dorsal interosseous (FDI) muscle of the right hand with Ag/AgCl electrodes, using a tendon-belly montage with an interelectrode distance of ±3 cm. The signal was sampled at 2048 Hz, band pass filtered ( Hz) and displayed in real time on a computer monitor for evaluation of EMG responses during the TMS procedures. EMG data were stored for off-line analysis of silent period durations without filtering. TMS procedure TMS pulses were delivered using a Magstim Rapid stimulator (Magstim, Whitland, UK) with a 70 mm figure-of-eight coil. The coil was placed over the contralateral left hemisphere tangential to the skull, with the handle pointing 45 posterolaterally. Prior to each rtms session, the motor hot spot was identified as the point where stimulation elicited the highest motor evoked potentials (MEPs). It was determined by moving the coil in 0.5 cm steps around C3 (according to the International system for EEG electrodes) using suprathreshold stimuli. After the position of the hot spot was established the stimulation intensity was decreased in steps of 1% to determine the resting motor threshold (RMT). RMT was defined as the intensity where MEPs of 50 lv or above were observed in at least 50% of 8

3 240 consecutive stimulations. rtms was delivered over the hand motor area and over the dorsal premotor cortex. Dorsal premotor cortex was identified as the location 2.5 cm anterior of the motor hand area. The rtms was delivered following current safety recommendations (Wasserman 1998). The experiment was performed in two different sessions, first for the motor and then the premotor cortex, approximately one month apart. During each session, trains of 1,500 stimuli at a frequency of 1 Hz were delivered at an intensity of 90% RMT (Fig. 1). In participants with relatively high RMTs, the coil was replaced during the trial when it overheated. Coil replacement did not exceed one minute. Before and after rtms application, the silent period (SP) was measured, which is an electrical silence interrupting the ongoing EMG activity of an actively contracting muscle following cortical stimulation. SP measures were obtained at an intensity of 110% and 120% of the RMT during FDI muscle contraction at 20% of maximum voluntary contraction. The SP duration was measured in the rectified EMG trace, from the end of the MEP elicited by the suprathreshold pulse, to the first point where EMG amplitude exceeded 50 lv. Sensorimotor synchronization task Participants sat at a table and were asked to tap on the surface of a force transducer (Novatech F241) with their right index finger in synchrony with the tones of an auditory metronome. Force was recorded at a sampling frequency of 1 khz. Tones were presented binaurally through headphones with an interstimulus interval of 500 ms. The sound of the finger touching the response plate was audible. Tapping was performed in fourteen blocks. Each block started with 10 tones in order to establish a baseline, followed by 8 consecutive trials of 24 tones each. A perturbation (phase shift) was introduced between the 7th and the 18th tone, resulting in a phase Fig. 1 Schematic of the experimental procedure. Participants first performed the sensorimotor synchronization task. Then, RMT was determined and SPs were measured with stimulation intensities of 110% and 120% RMT (black box). Low frequency subthreshold rtms was then applied over the motor hand area or the dorsal premotor cortex (grey box). At the end of stimulation, the RMT was determined again and the same measures of SP were obtained (black box). Immediately thereafter, participants performed the sensorimotor synchronization task again advance ( 10 or 50 ms) or a phase delay (+10 or +50 ms) of the remaining stimulus train. The magnitude and direction of the perturbations, as well as the position of the shift in the sequence, were presented in random order. Overall, each phase shift was presented 28 times. The duration of the tapping session was approximately 26 min. Tapping sessions were performed before and immediately after rtms. Data analysis The tap onsets were identified from the differentiated analogue force-time waveform, and the tone onsets from the analogue waveform of the tones, using a fixed threshold detection algorithm. Then, interresponse intervals (IRIs) and tap-tone asynchronies were computed. Asynchronies were computed by subtracting the tone onsets from the tap onsets and were positive if the tap occurred after the tone and negative if the tap preceded the tone. The mean of 6 tap-tone asynchronies before each phase shift (t-6 to t-1) was defined as the baseline asynchrony. Then, a time series including 6 asynchronies before (baseline) and 6 after the phase shift was extracted, to assess the way the asynchronies returned to the baseline (compensation function). In order to compare the compensation functions, the baseline asynchrony in each time series was subtracted, and the data were normalized for direction, by removing the sign in the phase-delayed conditions (+10, +50). This method allowed us to analyze the effects of phase advance and delay in the same analysis of variance. Statistical comparisons were performed separately for the two stimulation sites. Results Motor thresholds and silent periods The RMTs were not affected by stimulation of the motor or the premotor cortex. The mean (±SD) RMT before stimulation of the motor cortex was 53±8% and after 53±10%, and in the premotor cortex before 50±7% and after 50±6%. Durations of the SP before and after rtms for both motor and premotor cortices are summarized in Fig. 2. SPs were significantly longer after rtms application over both motor (t(7)=2.89, P<0.05) and premotor cortex (t(5) = 3.08, P<0.05).

4 241 Duration (ms) * p<0.05 Silent period motor cortex premotor cortex Negative asynchrony and error correction * pre post Fig. 2 Duration of the silent period pre and post rtms over the motor and premotor cortex. Vertical bars show one standard error of the mean All participants tapped with an anticipation tendency, yielding a negative asynchrony between the tap and the Fig. 3 Compensation functions for all experimental conditions before (filled circles) and after (open circles) motor cortex rtms. Increases (b, d) or decreases (a, c) of50(a, b) and 10 (c, d) ms in the metronome interval at response t result in rapidly corrected positive or negative shifts in the asynchrony between metronome pulse and tap. Vertical bars show one standard error of the mean * tone. Negative asynchrony values for performance before and after motor cortex stimulation are shown in Fig. 3. After motor cortex stimulation, a change in negative asynchrony was observed, indicating that participants were tapping closer to the tone (t(7)=2.768, P<0.05). Such a change in negative asynchrony was not found after premotor cortex stimulation (Fig. 4). A repeated measures analysis of variance (ANOVA) including data from participants that took part in both premotor and motor cortex stimulation (n=6) was performed with factors Stimulation (pre TMS, post TMS) and Site (motor, premotor). A significant interaction (F(1,5) = 7.8, P<0.05) confirmed that participants were tapping closer to the tone only after rtms over the motor cortex. No differences were found between pre and post stimulation measures of the asynchrony variance, or for the mean and variance of the IRIs. This was the case both for the motor (IRI pre: mean=499 ms, variance=994 ms 2, post: mean=500 ms, variance=977 ms 2 ) and the premotor cortex (IRI pre: mean=500 ms, variance=751 ms 2, post: mean=500 ms, variance=690 ms 2 ) stimulation. To examine whether the effect of stimulation over the motor cortex on the negative asynchrony was the same throughout the post stimulation period, the 112 sequences of each tapping session were split into 4 groups (group 1: 1 28, group 2: 29 56, group 3: 57 84, group 4: ), and mean asynchrony was calculated for each group. Repeated measures ANOVAs with factors Stimulation (pre TMS, post TMS) and Group (1, 2, 3, 4) revealed a main effect of Stimulation (F(1,7)=22.44, P<0.01), showing that participants were tapping closer to the tone after rtms, and this effect persisted a 0-20 Pre rtms Post rtms c Asynchrony (ms) b -40 d Asynchrony (ms) t-6 t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5 t+6 t-6 t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5 t+6

5 242 a 0-20 Pre rtms Post rtms c Asynchrony (ms) b d -60 Asynchrony (ms) t-6 t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5 t+6 t-6 t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5 t+6 Fig. 4 Compensation functions for all experimental conditions before (filled circles) and after (open circles) premotor cortex rtms associated with increases (b, d) or decreases (a, c)of50(a, b) and 10 (c, d) ms in the metronome interval at response t. Vertical bars show one standard error of the mean comparing impact force pre and post rtms showed no differences after motor (t(7)<1; pre TMS: 4.1±2.9 N; post TMS: 4.6±3.4 N) and premotor cortex stimulation (t(5)<1; pre TMS: 5.7±3.2 N post TMS: 6.1±3.4 N). throughout the post-stimulation period. No other reliable main effects or interactions were observed. Error correction was analyzed separately for motor and premotor cortex and for shift magnitude (50 and 10 ms). Three-way repeated measures ANOVAs with factors Stimulation (pre TMS, post TMS), Direction (phase advance, phase delay), and Position (t, t+1, t+2) were performed. A main effect of position was observed in all analyses (motor cortex 50 ms: F(2,6)=225.74, P<0.01, motor cortex 10 ms: F(2,6)=35.40, P<0.01, premotor cortex 50 ms: F(2,4)=103.90, P<0.01, premotor cortex 10 ms: F(2,4)=56.92, P<0.05) showing that the asynchrony was high at the point of the phase shift, and after a correction in the subsequent two responses it returned to the baseline (Fig. 4). There was neither a main effect of Stimulation (motor cortex 10 and 50 ms: F(2,6)<1, P>0.05, premotor cortex 10 and 50 ms: F(2,4)<1, P>0.05) nor any significant interaction involving this factor (motor cortex 10 and 50 ms: F(2,6)<1.5, P>0.05, premotor cortex 10 and 50 ms: F(2,4)<3, P>0.05). Impact force Maximum impact force was evaluated to investigate whether stimulation affected tapping force. T-tests Discussion The simple task of tapping to a metronome provides information on timing and the temporal coordination of actions with sensory events. Averaged over large numbers of trials, there is a close match of intertap intervals and metronome period, although taps usually slightly precede the metronome tones (Aschersleben and Prinz 1995). Remarkable timing precision is also demonstrated when regular metronome sequences are perturbed by temporal irregularities, such as the metronome phase shifts used in the present study. Phase shifts as small as 5 15 ms, which cannot be consciously perceived, have previously been shown to be corrected within a few taps (e.g. Repp 2000; Praamstra et al. 2003). One important result of the present study is that this rapid adjustment to perturbations is unaffected by rtms applied to motor or lateral premotor cortex. Preserved correction of the larger magnitude 50 ms phase shifts is unsurprising. These larger shifts were still unobtrusive but nonetheless easy to perceive. Conscious perception of the shifts means that central timing mechanisms could be invoked to support the adjustment (Repp 2001a; Praamstra et al. 2003). We did not anticipate that these central timing mechanisms, probably involving the supplementary motor area (SMA) (Rao et al. 1999; Lewis et al. 2004; Macar et al. 2002), would

6 243 be affected by rtms applied to motor or premotor cortex. Correction of the smaller perturbations, by contrast, is an automatic process that might be dependent on structures within the CNS closer to the motor output level, such as the sensorimotor cortex. Hence, stimulation of motor and premotor cortex was expected to interfere with the correction of the 10 ms phase shifts. The finding that correction of the small magnitude 10 ms phase shifts was unaffected by TMS, combined with the absence of any effect on the variability of tapping, seems to suggest that motor and premotor cortex are not critically involved in the ability to synchronize to a metronome. However, our reasoning so far does not yet take into account all the experimental findings. Preserved correction of stimulus phase shifts contrasted with a sustained shift in the tap-tone asynchrony. That is, with stimulation applied to the motor cortex, but not when applied to the premotor cortex, participants tapped closer to the tone, with a reduction of the negative asynchrony of approximately 10 ms. This latter finding indicates that rtms of the motor cortex, with stimulation parameters inducing an increased inhibition of the motor cortex, does influence sensorimotor synchronization. In the following we will discuss several lines of explanation that might account for the altered tap-tone asynchrony with preserved adjustment to metronome stimulus phase shifts. A change in tap-tone asynchrony without any effect on the ability to correct timing perturbations looks contradictory at first sight. Experimental designs with timing perturbations are used precisely to investigate the correction mechanisms that enable participants to tap to an isochronous metronome without the tapping drifting away from the stimulus sequence (Hary and Moore 1987; Repp 2000; Vorberg and Wing 1996). One line of explanation to account for an effect on asynchrony but not on the correction process derives from the work of Vorberg and Wing (1996), who suggested that onset asynchrony can be interpreted as due to an error in timekeeper period. Provided there is continual phase correction, stable phase with a pacing stimulus can be maintained, even when the period is different from the metronome interval. If the period is shorter than the metronome, the asynchrony is negative, if it is longer the asynchrony becomes positive. In our case, inhibition of the sensorimotor cortex by rtms may have resulted in a change in the timekeeper period relative to the metronome interval. For instance, if the timekeeper is based on an underlying pulse-counting process (Gibbon et al. 1984) and the pulse is slowed by rtms, for any given target count, the resulting interval will be lengthened. The reduction in negative asynchrony in our study then reflects a reduction in the normal shortening of the timekeeper period so that it actually becomes more nearly equivalent to the metronome interval. A recent study investigated rtms over dorsolateral prefrontal cortex, an area associated with working memory and attention (Jones et al. 2004). In contrast with our proposed account, rtms shortened the reproduction of time intervals, but only when the interval was in the seconds rather than milliseconds range. This finding suggests a contrast between neural circuitry supporting short and long interval production. Such a contrast is consistent with Lewis and Miall (2003) who argued, on the basis of neuroimaging data, that timing in the milliseconds range of intervals is a function of motor circuits, whereas timing in the seconds range is associated with cognitive functions of the prefrontal cortex. However, in our account it is unclear why rtms over premotor cortex would not have resulted in a similar effect as motor cortex stimulation. A second line of explanation is based on recent magnetoencephalography (MEG) investigations of sensorimotor synchronization by Mu ller et al. (2000). These authors identified MEG activity originating from inferior parts of the primary sensorimotor cortex (SI) which they associated with the evaluation of the asynchrony of tap and auditory metronome stimuli. Conceivably, our stimulation of motor cortex also affected post-central sensory cortex, influencing the putative inferior SI area where auditory and somatosensory information is integrated. If the effect had been to advance the registration of the tap or delay the registration of the auditory event, the asynchrony would have been reduced without affecting the compensation function. The absence of an effect of premotor cortex rtms would be interpreted as a lack of effect on sensory cortex with the more anterior positioning of the stimulation. However, an explanation along these lines appears suspect because rtms effects on somatosensory processing might have been expected to be suppressive rather than facilitatory (Enomoto et al. 2001) and a more recent MEG study of sensorimotor synchronization failed to replicate the finding of integrative activity in inferior SI (Pollok et al. 2004). The above explanation, with a putative brain area that is sensitive to the temporal asynchrony of tap and tone, fits with formal models of error correction in sensorimotor synchronization that take the tap-tone asynchrony as the critical information on which error correction is based (Mates 1994; Vorberg and Wing 1996). However, this assumption might be questioned. The observation that participants rapidly correct small metronome phase shifts, even when they do not perceive them, has been interpreted as an example of very sensitive processes in support of perception for action (Repp 2000, 2001a; Praamstra et al. 2003), reminiscent of automatic corrections in the visuomotor domain (Milner and Goodale 1995). A model of sensorimotor synchronization that avoids the assumption of such sensitive asynchrony processing is the so-called mixed phase-resetting model (Hary and Moore 1985; 1987; Repp submitted), which does not regard the taptone asynchrony as the informational basis for phase correction. Instead, it proposes that during sensori-

7 244 motor synchronization individual taps will be timed relative to the preceding tap or relative to the preceding metronome stimulus, with the choice between the two temporal reference points being random or determined by a dynamic competition (Repp submitted). How is the mixed phase-resetting model of error correction in sensorimotor synchronization relevant to our rtms findings? We already considered the possibility of sensory effects when stimulation is applied to the motor cortex. Effects on somatosensory function have been found in a study using stimulation parameters very similar to ours (Enomoto et al. 2001). These effects were suppressive in nature, as demonstrated by means of somatosensory evoked potential (SEP) amplitude reductions, which moreover occurred only with stimulation of MI but not with stimulation at sites overlying premotor or sensory cortex. We hypothesize that inhibitory effects on SI attenuate or change the sensory perception of the tap in a way that gives more prominence to the metronome stimulus as a later temporal reference point for period generation. The tighter coupling of taps and tones, i.e. the reduction of the anticipation tendency, is thus seen as the consequence of a change in balance between the temporal reference point corresponding to the tap and the reference point corresponding to the tone. This line of explanation might be seen as being contradicted by evidence from sensory deafferented patients, who have an enhanced anticipation tendency (Aschersleben et al. 2002), whereas our argument would suggest there should be less anticipation with reduced somatosensory input from each tap. However, it could be argued that, in the absence of sensory afference, there are strategic changes in task performance which could result in quite different overall effects from those induced by TMS. Another difficulty of the proposed explanation is that it would imply faster phase correction, an effect not observed in our results. Possibly, faster correction was not observed because of a ceiling effect, i.e. the correction was already very swift with a return to the baseline asynchrony within two responses. A remaining possibility that has to be considered is that the altered sensory feedback made subjects attend closer to the auditory feedback provided by the audible collision of the finger on the force plate. Although subjects wore headphones, the sounds of the collision were not completely suppressed. Greater attention to auditory feedback could result in a reduction of the negative asynchrony because it enhances perception of asynchronies within the auditory modality (Aschersleben and Prinz 1995). We hypothesized that with greater reliance on the auditory feedback, subjects would tap with more force after TMS to make the taps better audible. However, no such effect was observed, as reported above. Note further that with this explanation the shortening of the negative asynchrony should be accompanied by faster error correction, which was not the case. Taken together, there is no specific evidence to support this explanation, although it cannot be ruled out. Conclusion We found differential effects of premotor cortex and motor cortex stimulation on sensorimotor synchronization. Stimulation of both areas increased motor cortex inhibition, but only motor cortex stimulation had an influence on sensorimotor performance. This dissociation suggests that rtms had an effect on the somatosensory cortex or, alternatively, reduced the sensitivity of the primary motor cortex to somatosensory input. The latter possibility is in agreement with previous rtms and SEP studies (Enomoto et al. 2001; Lee et al. 2003) and implies that premotor and motor cortex stimulation have different effects on sensory input to the motor cortex, in spite of similar reductions of corticospinal excitability. Effects on motor performance were also characterized by a dissociation, i.e. an unaffected correction of induced phase shifts contrasting with a shift in the baseline tap-tone asynchrony. Although this pattern of results is not readily explained within current models of sensorimotor synchronization, we have outlined several possible lines of explanation to accommodate this observation. The results support a role of sensorimotor cortex in synchronization performance and single out this task as one of few simple motor behaviours affected by rtms induced motor cortex inhibition. References Aschersleben G, Prinz W (1995) Synchronizing actions with events: the role of sensory information. Percept Psychophys 57: Aschersleben G, Stenneken P, Cole J, Prinz W (2002) Timing mechanisms in sensorimotor synchronization. In: Prinz W, Hommel B (eds) Common mechanisms in perception and action: attention and performance, vol XIX. Oxford University Press, Oxford, pp Chen R, Classen J, Gerloff C, Celnik P, Wassermann EM, Hallett M, Cohen LG (1997) Depression of motor cortex excitability by low-frequency transcranial magnetic stimulation. Neurology 48: Enomoto H, Ugawa Y, Hanajima R, Yuasa K, Mochizuki H, Terao Y, Shiio Y, Furubayashi T, Iwata NK, Kanazawa I (2001) Decreased sensory cortical excitability after 1 Hz rtms over the ipsilateral primary motor cortex. Clin Neurophysiol 112: Fraisse, P (1980) Les synchronizations sensori-motrices aux rythmes. [Sensorimotor synchronization to rhythms]. In J Requin (ed) Anticipation et comportement. Centre National, Paris, pp Gerschlager W, Siebner HR, Rothwell JC (2001) Decreased corticospinal excitability after subthreshold 1 Hz rtms over lateral premotor cortex. Neurology 57: Gibbon J, Church RM, Meck WH (1984) Scalar timing in memory. Ann N Y Acad Sci 423:52 77 Hary D, Moore GP (1985) Temporal tracking and synchronization strategies. Hum Neurobiol 4:73 79 Hary D, Moore GP (1987) Synchronizing human movement with an external clock source. Biol Cybern 56: Jäncke L, Loose R, Lutz K, Specht K, Shah NJ (2000) Cortical activations during paced finger-tapping applying visual and auditory pacing stimuli. Brain Res Cogn Brain Res 10:51 66

8 245 Jäncke L, Steinmetz H, Benilow S, Ziemann U (2004) Slowing fastest finger movements of the dominant hand with low-frequency rtms of the hand area of the primary motor cortex. Exp Brain Res 155: Jones CR, Rosenkranz K, Rothwell JC, Jahanshahi M (2004) The right dorsolateral prefrontal cortex is essential in time reproduction: an investigation with repetitive transcranial magnetic stimulation. Exp Brain Res 158: Khedr EM, Gilio F, Rothwell J (2004) Effects of low frequency and low intensity repetitive paired pulse stimulation of the primary motor cortex. Clin Neurophysiol 115: Kobayashi M, Hutchinson S, Theoret H, Schlaug G, Pascual- Leone A (2004) Repetitive TMS of the motor cortex improves ipsilateral sequential simple finger movements. Neurology 62:91 98 Lee L, Siebner HR, Rowe JB, Rizzo V, Rothwell JC, Frackowiak RS, Friston KJ (2003) Acute remapping within the motor system induced by low-frequency repetitive transcranial magnetic stimulation. J Neurosci 15: Lewis PA, Miall RC (2003) Distinct systems for automatic and cognitively controlled time measurement: evidence from neuroimaging. Curr Opin Neurobiol 13: Lewis PA, Wing AM, Pope PA, Praamstra P, Miall RC (2004) Brain activity correlates differentially with increasing temporal complexity of rhythms during initialisation, synchronization, and continuation phases of paced finger tapping. Neuropsychologia 42: Macar F, Lejeune H, Bonnet M, Ferrara A, Pouthas V, Vidal F, Maquet P (2002) Activation of the supplementary motor area and of attentional networks during temporal processing. Exp Brain Res 142: Maeda F, Keenan JP, Tormos JM, Topka H, Pascual-Leone A (2000) Modulation of corticospinal excitability by repetitive transcranial magnetic stimulation. Clin Neurophysiol 111: Mates J (1994) A model of synchronization of motor acts to a stimulus sequence I Timing and error corrections. Biol Cybern 70: Milner AD, Goodale MA (1995) The Visual Brain in Action. Oxford University Press, Oxford Modugno N, Curra A, Conte A, Inghilleri M, Fofi L, Agostino R, Manfredi M, Berardelli A (2003) Depressed intracortical inhibition after long trains of subthreshold repetitive magnetic stimuli at low frequency. Clin Neurophysiol 114: Muellbacher W, Ziemann U, Boroojerdi B, Hallett M (2000) Effects of low-frequency transcranial magnetic stimulation on motor excitability and basic motor behavior. Clin Neurophysiol 111: Mu ller K, Schmitz F, Schnitzler A, Freund H-J, Aschersleben G, Prinz W (2000) Neuromagnetic correlates of sensorimotor synchronization. J Cogn Neurosci 12: Mu nchau A, Bloem BR, Irlbacher K, Trimble MR, Rothwell JC (2002) Functional connectivity of human premotor and motor cortex explored with repetitive transcranial magnetic stimulation. J Neurosci 22: Paillard J (1948) Quelques données psychophysiologiques relatives au de clenchement de la commande motrice [Some psychophysiological data relating to the triggering of motor commands]. L Année Psychologique 47 48, Pascual-Leone A, Tormos JM, Keenan J, Tarazona F, Canete C, Catala MD (1998) Study and modulation of human cortical excitability with transcranial magnetic stimulation. J Clin Neurophysiol 1: Pollok B, Mu ller K, Aschersleben G, Schnitzler A, Prinz W (2004) The role of the primary somatosensory cortex in an auditorily paced finger tapping task. Exp Brain Res 156: Praamstra P, Turgeon M, Hesse CW, Wing AM, Perryer L (2003) Neurophysiological correlates of error correction in sensorimotor synchronization. Neuroimage 20: Rao SM, Harrington DL, Haaland KY, Bobholz JA, Cox RW, Binder JR (1997) Distributed neural systems underlying the timing of movements. J Neurosci 15: Repp BH (2000) Compensation for subliminal timing perturbations in perceptual-motor synchronization. Psychol Res 63: Repp BH (2001a) Phase correction, phase resetting, and phase shifts after subliminal timing perturbations in sensorimotor synchronization. J Exp Psychol Hum Percept Perform 27: Repp BH (2001b) Processes underlying adaptation to tempo changes in sensorimotor synchronization. Hum Mov Sci 20: Repp BH, Keller PE (2004) Adaptation to tempo changes in sensorimotor synchronization: effects of intention, attention, and awareness. Q J Exp Psychol 57: Repp BH (2003) Rate limits in sensorimotor synchronization with auditory and visual sequences: The synchronization threshold and the benefits and costs of interval subdivision. J Mot Behav 35: Ridding MC, Taylor JL, Rothwell JC (1995) The effect of voluntary contraction on cortico-cortical inhibition in human motor cortex. J Physiol 487: Rizzo V, Siebner HR, Modugno N, Pesenti A, Mu nchau A, Gerschlager W, Webb RM, Rothwell JC (2004) Shaping the excitability of human motor cortex with premotor rtms. J Physiol 15: Rossi S, Pasqualetti P, Rossini PM, Feige B, Ulivelli M, Glocker FX, Battistini N, Lucking CH, Kristeva-Feige R (2000) Effects of repetitive transcranial magnetic stimulation on movementrelated cortical activity in humans. Cereb Cortex 10: Semjen A, Schulze HH, Vorberg D (2000) Timing precision in continuation and synchronization tapping. Psychol Res 63: Siebner HR, Rothwell J (2003) Transcranial magnetic stimulation: new insights into representational cortical plasticity. Exp Brain Res 148:1 16 Sommer M, Wu T, Tergau F, Paulus W (2002) Intra- and interindividual variability of motor responses to repetitive transcranial magnetic stimulation. Clin Neurophysiol 113: Stephan KM, Thaut MH, Wunderlich G, Schicks W, Tian B, Tellmann L, Schmitz T, Herzog H, McIntosh GC, Seitz RJ, Homberg V (2002) Conscious and subconscious sensorimotor synchronization prefrontal cortex and the influence of awareness. Neuroimage 15: Theoret H, Haque J, Pascual-Leone A (2001) Increased variability of paced finger tapping accuracy following repetitive magnetic stimulation of the cerebellum in humans. Neurosci Lett 22:29 32 Vorberg D, Wing A (1996) Modelling variability and dependence in timing. In: Heuer H, Keele SW (eds) Handbook of perception and action. Motor skills, vol 2. Academic Press, London, pp Wassermann EM (1998) Risk and safety of repetitive transcranial magnetic stimulation: report and suggested guidelines from the International Workshop on the Safety of Repetitive Transcranial Magnetic Stimulation, June 5 7, Electroencephalogr Clin Neurophysiol 108:1 16