HRV has been found useful in the study of cardiological illness in adults and elders, as well as in monitoring prenatal health. Twenty-four hour Holter recordings of R to R intervals (RRI) in healthy newborns, adults, and elderly persons were analyzed with statistical, chaos, and recurrence methods. In persons of all ages, RRI series showed relative stability (as expected in homeostatic regulation), patterned daily changes in heart rate, evidence of causality or 'determinism' (nonrandom pattern of the series of differences), and non-periodic irregular variations within limits, suggesting chaos. In addition, novel methods of analysis reveal creative features that are absent in chaotic attractors but found in bios, a non-stationary process that is generated mathematically by recursions of bipolar feedback (chaotic bios) or by the addition of sine waves. Wavelet and recurrence plots demonstrate time-limited patterns (e.g. clustering of recurrences in organized complexes) that follow each other in time indicating temporal complexity, in contrast to the temporal uniformity of chaotic attractors and of random changes. Recurrence quantification demonstrates less recurrence isometry than copies randomized by shuffling (novelty), and more consecutive isometries than shuffled copies indicating causal order. Statistical analyses demonstrate asymmetric distribution and diversification (increase in variance with the duration of the series analyzed) in contrast to convergence to an attractor. These studies indicate that the normal pattern of HRV is both homeostatic and biotic. A biotic pattern with homeostatic features (homeobios) is generated by combining bipolar feedback with negative feedback. Chaos and bios analyses may thus be useful in clinical studies.