Respiratory sinus arrhythmia (RSA)
The near-universal non-invasive index of cardiac parasympathetic control, and the single measure on which almost all vagal psychophysiology — including all of Polyvagal Theory — depends. Enters the wiki through the Porges/Grossman debate, where its meaning is precisely what is contested.
What it is
Heart rate is not metronomic: it speeds slightly on inhalation and slows on exhalation. The amplitude of this respiration-linked fluctuation is RSA, essentially equivalent to high-frequency heart-rate variability (HF-HRV) in the respiratory band. Because this modulation is carried by the vagus, RSA is read as a window onto cardiac vagal tone — the vagal contribution to resting heart rate. Higher RSA generally tracks better health, resilience and regulation (a large literature both sides accept).
The distinction the debate turns on
RSA (an index), cardiac vagal tone (the vagal control of the heart it approximates), and general/organ-wide vagal tone are three different things, and the wiki keeps them apart because the polyvagal dispute is exactly a dispute about collapsing them:
- Porges treats RSA amplitude as “a validated and easily obtainable index of parasympathetic nervous system tone via the cardiac vagus,” and specifically as a readout of the ventral (nucleus ambiguus) vagal brake — the portal through which polyvagal hypotheses are tested. He argues his Porges–Bohrer (PBRSA) quantification isolates the ventral-vagal signal.
- Grossman calls the equation RSA = cardiac vagal tone (still more RSA = general vagal tone) a Rylean category mistake. RSA is only an approximate index of cardiac vagal tone, and only when respiration is controlled (a requirement since Katona & Jih 1975); it says nothing about vagal outflow to the lungs or gut, because of vagal target-organ specificity (cardiac, pulmonary and gastric vagal activity do not covary — see autonomic-specificity). He reports PBRSA is not superior to other metrics and is itself respiration-dependent (r ≈ 0.84 with respiratory rate).
Two consequences the wiki should hold onto regardless of who wins:
- RSA is respiration-contaminated. Anything that changes breathing rate or depth changes RSA, whether or not cardiac vagal tone changed. This is why emotion→RSA effects can be a breathing artefact (respiratory-interoception).
- A cardiac index cannot speak for “the vagus.” The step from “RSA changed” to “vagal tone changed” to “the autonomic state changed” is three inferential jumps, each contestable.
Where it sits in the wiki
RSA is the measurement backbone of polyvagal-theory and the evidence base of its clinical uses; it recurs across the trauma and intervention literature (Weng reports RSA rising after MABT; SE leans on ventral-vagal tone). It is a physiological cousin of the wiki’s other cardiac measures but measures something different from heartbeat perception (heartbeat-detection-task, heart-rate-discrimination-task): RSA is an efferent-autonomic readout requiring no awareness, not an interoceptive-accuracy score. Its predictive validity for behaviour and adaptation is the subject of a meta-analytic literature (three RSA meta-analyses were queued in raw/; all three are now in). All three landed results are cautionary. Wesarg et al. (2022): pooling ~27,000 participants, childhood adversity — one of the field’s most robust developmental risk factors — shows no overall association with either baseline vagal activity or vagal reactivity, and only a small one (r ≈ −.05 to −.11) in clinical samples and for direct/maltreatment adversity. Graziano & Derefinko (2013): across 44 studies and ~4996 children, RSA withdrawal predicts fewer externalizing, internalizing and cognitive/academic problems (each r ≈ −.16, small) but is unrelated to social functioning (r = −.06) — the very domain Polyvagal Theory’s Social Engagement System is built to explain, and the effect even flips maladaptive in clinical samples. Zahn et al. (2016): across 26 studies, resting HRV predicts laboratory self-control at only r = 0.15, and that small effect drops to a nonsignificant r = 0.04 once publication bias is corrected — the Neurovisceral-Integration prediction (HRV as prefrontal inhibitory capacity) barely survives scrutiny. Whatever RSA indexes, its individual-difference traffic with real-world outcomes is thinner than the theories built on it assume — the same deflation the reliability and domain-generality literatures reach for other body-based measures.