Atrial fibrillation (AF) is one of the most common cardiac arrhythmias, affecting around 10% of the population older than 70 years. In AF, the normal electrical impulses that are generated by the sinoatrial node are overwhelmed by disorganized electrical impulses within the atria. This is associated with an irregular bombarding of the atrial activity into the AV node. Since the AV node is not able to conduct all atrial activations, some of them are blocked within the node. This "filtering" property of the AV node is fundamental to maintain the heart beating in a range compatible with the life. However, the ventricular responses during AF presents shorter and more irregular RR intervals (time between two heart beats) than during normal sinus rhythm. This fast and irregular ventricular rate associated can cause some of the most severe symptoms of this arrhythmia. Since the AV node is the only normal structure responsible for the conduction of atrial impulses to the ventricles, the strategy of control the heart rate during AF essentially deals with efforts to utilize and adjust conduction properties of the node. However, the role of AV nodal conduction properties in controlling and modulating the ventricular response during AF is not well understood.

During the development of the present thesis different characteristics of the AV conduction have been investigated in different species and with different techniques with the aim throw light upon this intriguing structure of the heart. Specifically, we analyzed one of the most intriguing behaviors of the ventricular response patterns during AF; when constructing RR-interval histograms obtained from Holter recordings with persistent AF, uni-, bi-, or multimodal RR distribution patterns can be found. These predominant RR intervals have been suggested in the literature to be multiples of the refractory period of the AV node or caused by the existence of a dual AV node physiology. However, during the development of this thesis some results incompatible with these theories have been presented, wherefore we present and defend a novel hypothesis suggesting that predominant RR intervals are related to the atrial fibrillatory process. In order to validate this hypothesis we have used two main approaches, one to confirm empirically our assumptions by using real human clinical recordings and another to validate those assumptions by using mathematical models of the atrial fibrillatory process and the atrioventricular conduction.

In the first part of the present dissertation, Holter ECG signals from patients with persistent AF were analyzed. Number and position of predominant RR intervals were detected and compared with mean and standard deviation of the dominant atrial cycle length. It allowed us to demonstrate a statistical relation between the position of predominant RR intervals multiples of the dominant atrial cycle length.

In the second part of this thesis, epicardial recordings from humans and rabbits were used to mathematically model the atrial fibrillatory process and the atrioventricular conduction. By using these mathematical models the fundamental role of the mean AA interval series has been illustrated; the AV conduction during AF has been demonstrated to be confined to scaling the mean atrial rate. The importance of the organization of the atrial bombarding into the AV node has been studied. In added, developed mathematical models were also useful to elucidate some complex peculiarities of conduction through the AV node during arrhythmias. The model has been applied to explain how the interaction between FP and SP propagation may be responsible for the development of atypical Wenckebach rhythms (irregular periodic ventricular cycles during regular atrial activation). In addition, the presented model has allowed an easy modification of either FP or SP in the same heart thus permitting evaluation of the effects of clinical ablations used as therapeutic tools for control of ventricular rate during AF.