Bases neurales y comportamentales de la preparación mediante ritmos

Résumé

Temporal regularities present in our environment allow us not only to synchronize our movements to an external rhythm but also to generate temporal expectations about when a relevant event is going to occur (Nobre, Correa, & Coull, 2007). This synchronization, known as rhythmic entrainment, results in behavioural benefits in response to the events matching the temporal structure of the sequences, such as an enhancement of the reaction times (RT; Correa & Nobre, 2008; Sanabria, Capizzi, & Correa, 2011; Sanabria & Correa, 2013), time judgment (Barnes & Jones, 2000) or pitch judgment accuracy (Jones, Moynihan, McKenzie, & Puente, 2002). While several studies have suggested that this process is independent of top-down attention (Rohenkohl, Coull, & Nobre, 2011; Triviño et al., 2011; Correa et al., 2014), to our knowledge, only two of them have focused directly on the role of attention in the rhythmic entrainment, showing contradictory results (De la Rosa, Sanabria, Capizzi, & Correa, 2012; Schwartze, Rothermich, Schmidt-Kassow, & Kotz, 2011). The aim of the present thesis is to investigate whether top-down attentional processes are involved in the rhythmic entrainment and, moreover, to account for the role of the rhythmic entrainment as a basic process in a general predictive system. De la Rosa and colleagues (De la Rosa, Sanabria, Capizzi, & Correa, 2012) conducted an experiment under the dual-task paradigm in order to study the resistance of rhythmic entrainment to the concurrent performance of a secondary working memory task. According to the dual-task paradigm, the inclusion of a secondary would result in an impairment of performance in the primary task if both tasks compete for common limited resources (Logan, 1978, 1979). De la Rosa and colleagues (2012, Experiment 2) presented auditory isochronous (the interval between the stimuli was always the same) and anisochronous sequences (the interval between the stimuli was randomly distributed). The task of the participants was to respond as fast as possible to a target presented after these sequences. An adaptation of the Sternberg WM task (Sternberg, 1966) had to be performed concurrently with the rhythmic entrainment task. They asked participants to remember a series of either six different letters (high-load condition) or only one letter presented six times (low-load condition) that were visually presented at the beginning of every trial. Finally, after the target onset, a random letter appeared on the screen and participants had to respond whether or not this letter was among the ones presented at the beginning of the trial. Results of De la Rosa et al.’s (2012) study showed that the entrainment effect (faster RT after isochronous than after anisochronous sequences) was not affected by the inclusion of the secondary task. These results suggest and independence of the rhythmic entrainment from top-down attentional processes. However, it has been suggested that the competition for cognitive resources is reduced when two tasks are presented in different sensory modality (Wickens, 1980, 1984). In the Experimental Series I of the present thesis we wanted to address the issue of the sensory modality of the two competing tasks by presenting the secondary working memory task within the same modality as the rhythmic entrainment task. In Experiment 1, we adapted the task designed by De la Rosa et al. (2012) by presenting both, the Sternberg type WM task and the rhythmic entrainment task, auditorily. Results showed that the rhythmic entrainment was resistant to the inclusion of a secondary task. However, there was not a significant main effect of the memory load on the RTs, since participants were equally faster regardless of the memory load condition and thus, it could be that the secondary task was not demanding enough to produce an interference in the rhythmic entrainment. To address this issue, in Experiment 2, we used an n-back task instead of the Sternberg type task used in the Experiment 1. It has been shown that tasks requiring memory updating of the information (e.g., n-back task) interfere in the temporal processing of intervals, while tasks requiring only maintenance of the information (e.g. Sternberg type) do not show this interference (Rammsayer & Ulrich, 2011). Moreover, memory updating has been proposed as an executive process (Miyake et al., 2000; Salmon et al., 1996). With this manipulation, results of Experiment 2 showed a significant main effect of the memory load, this is, slower RTs to the targets. However, this effect was presented only globally, since the difference in RT between the isochronous and anisochronous conditions remained the same regardless of the memory load. To sum up, these results showed that isochronous sequences can result in rhythmic entrainment regardless of the concurrent performance of a demanding secondary task presented within the same modality, suggesting, thus, that rhythmic entrainment is independent of top-down attention. The Experimental Series II was designed under the assumption that the involvement of top-down attention in rhythmic entrainment might be gradual, since it has been suggested that the automaticity of cognitive processes might be seen from a continuum instead of from a dichotomic perspective (MacLeod & Dunbar, 1988; Capizzi, Sanabria, & Correa, 2012). In Experiments 3-4, we manipulated the anisochrony of the sequences, resulting in different sequences ranging from smaller to larger deviations of the isochrony (0, 10, 20, 50, 100, 150, and 200 ms). This manipulation not only gave us the opportunity to study the gradation of the automaticity in the rhythmic entrainment, but also to observe a behavioural pattern of the rhythmic entrainment. In Experiment 3, we decided to use the manipulation of synchrony described above to test if the pattern of the auditory rhythmic entrainment was different to the pattern observed after presenting visual sequences. Although auditory temporal processing has shown an advantage over visual processing (Glenberg & Jona, 1991; Repp & Penel, 2002; Guttman, Gilroy, & Blake, 2005; Welch, Dutton Hurt, & Warren, 1986) it remains unclear whether this advantage could be extended to the rhythmic entrainment. With the purpose to clarify this issue, we presented the sequences in both auditory and visual conditions. Results of Experiment 3 showed an entrainment window in which participants were equally faster after isochronous sequences and also after smaller deviations of this isochrony (10, 20 and 50 ms deviations), suggesting that anisochronous sequences can also result in rhythmic entrainment to some extent. Moreover, the RTs were gradually slower as the deviations from the isochrony increased, showing a behavioural benefit of the sequences after the entrainment window when compared with larger deviations (i.e., faster RTs at the 100 ms condition than at the 200 ms). Importantly, no differences were found between the auditory and visual conditions. In Experiment 4, we included a n-back task that participants had to perform concurrently with the task described in Experiment 3. Once again, the entrainment effect was resistant to the inclusion of this task and the behavioural pattern in response to the target remained the same for both, auditory and visual modalities. Taking together, the results of Experiments 3-4 suggest that the rhythmic entrainment constitutes a flexible process and that its effect does not depend on the modality in which the sequences are presented. Finally, the lack of interference of the secondary task points to a complete independence between rhythmic entrainment and top-down attention. In the Experimental Series III, we went a step further to test role of attention in the rhythmic entrainment by including the recording of event-related potentials (ERPs) to study the phenomenon at hand. This study was of particular interest, since, as far as we know, the only study that suggests that rhythmic entrainment is dependent of topdown attention found an enhancement of the P3 component when deviant tones were presented within isochronous sequences compared to when the deviant tones were presented within anisochronous sequences. Importantly, this enhancement was only found when participants were explicitly instructed to attend to the stimuli (Schwartze, Rothermich, Schmidt-Kassow, & Kotz, 2011). In Experiment 5, we used the same task as in the Experiment 2 of this thesis, this is, two conditions for the rhythmic entrainment task (isochronous and anisochronous) and the inclusion of a n-back task as a secondary task with different memory loads (0-back and 2-back). Contrary to the results showed by Schwartze and colleagues (2011) we found an enhancement of the P3 component to the targets presented after anisochronous sequences regardless of the memory load condition, moreover, the amplitude of the P3 also showed significant differences between modalities in the working memory task (larger amplitude in the 2-back condition). The enhancement of the P3 component when unexpected or novel stimuli are detected (novelty P3, NP3; Friedman, Cycowicz, & Gaeta, 2001) can explain both, the enhancement after anisochronous sequences in RT tasks and the enhancement after the detection of deviants in isochronous sequences. Results of the Experiment 5 suggest, once again, that top-down attention is not necessary for the isochronous sequences to produce rhythmic entrainment. Up to this point, we have studied the automaticity of the rhythmic entrainment with manipulations of working memory tasks and of the entrainment sequences in both, behavioural and electroencephalographic studies. However, it could be argued that the resources necessary to perform working memory tasks are hardly related to the rhythmic entrainment. In the Experimental Series IV (Experiment 6), we addressed this issue by including a sensorymotor synchronization task as a secondary task. This task is of particular interest,