Abstract
Standard heart rate variability (HRV) metrics capture how far the autonomic system moves between states (excursion) but not in how many independent directions it moves (dimensionality). The Response Repertoire Index (RRI-D) quantifies dimensionality as a Hill number of order 2 of the eigenvalue spectrum of four HRV features measured across standardized states; a preregistered test in 41 healthy adults found it computable, acceptably reliable at group level and unrelated to excursion. This framework paper connects the construct to resilience research, where resilience is understood as a process of adaptation, and positions RRI-D as a candidate marker of regulatory latitude, not as a resilience measure. It states three falsifiable hypotheses: (H1) under sustained load RRI-D declines while resting RMSSD stays practically unchanged, a possible counterpart of critical slowing down; (H2) interventions acting on amplitude and interventions acting on multi-state regulation dissociate on excursion versus RRI-D; (H3) autonomic dimensionality and a psychological repertoire index (RRI-R) relate to strain and recovery in additive, multiplicative or threshold form. Simulations of the registered algorithm show that RRI-D orders true dimensionality at group level, that measurement noise inflates it at low excursion, and that the observed near-zero correlation with excursion is what independence plus noise predicts. They also give sample sizes: 36 persons with five sessions detect a within-person decline of 0.1 units with 84% power, whereas distinguishing the H3 models needs 200 to 400 persons. We propose a minimal multi-state protocol for data sharing. RRI-D is not validated for individual use, and no clinical or efficacy claim is made.