The purpose of this study was to expand classic heart period analysis methods by techniques from ethnomusicology that explicitly take complex musical rhythm principles into consideration. The methods used are based on the theory of African music, the theory of symbolic dynamics, and combinatorial theory. Heart period tachograms from 192 24-h electrocardiograms of 96 healthy subjects were transformed into binary symbol sequences that were interpretable as elementary rhythmic (percussive) patterns, the time lines in African music. Using a hierarchical rhythm pattern scheme closely related to the Derler Rhythm Classification (from jazz theory), we calculated the predominance and stability of pattern classes. The results show that during sleep certain classes, specific to individuals, occurred in a cyclically recurrent manner and many times more often than expected. Simultaneously, other classes disappeared more or less completely. Moreover, the most frequent classes obviously originate from phase-locking processes in autonomic regulation (e.g., between respiratory and cardiac cycles). In conclusion, the new interdisciplinary method presented here demonstrates that heart period patterns, in particular those occurring during night sleep, can be interpreted as musical rhythms. This method may be of great potential use in music therapy research.