Heart rate variability (HRV)

Method at a glance

• Modality: cardiac
• Measures: Beat-to-beat variation in the interval between heartbeats, recorded from ECG or PPG. Its vagally-mediated components (HF power, RMSSD, and respiratory sinus arrhythmia) are used as a non-invasive index of parasympathetic (cardiac vagal) influence on the heart. It is an efferent autonomic readout, not a measure of interoceptive perception.
• Procedure: Record inter-beat intervals (ECG or PPG) at rest, and optionally during and after a challenge task. Summarise with time-domain (RMSSD, SDNN) or frequency-domain (HF power) metrics, or with RSA-specific quantifications such as Porges-Bohrer or peak-to-trough. Trait studies use a resting baseline; reactivity studies compare task to baseline via task-mean levels, difference scores or residualized change. Respiration should be measured or controlled, since breathing rate and depth change the vagal components directly.
• Also called: HRV, vmHRV, vagally-mediated HRV, resting HRV, RMSSD, HF-HRV, SDNN

Strengths

• Non-invasive, cheap and easily collected alongside other measures (e.g. concurrent PPG during resting-state fMRI in Banellis et al. 2026), so it scales to large samples and meta-analysis.
• Requires no report or awareness from the participant, so it avoids the confidence and belief confounds that affect heartbeat-perception tasks.
• A large literature, accepted on both sides of the polyvagal dispute, links higher resting vagal HRV with better health, resilience and regulation.
• Serves as an objective physiological outcome for interventions (e.g. RSA improvement after MABT; HRV biofeedback and slow breathing).

Limitations

• Respiration-contaminated: the vagal components shift with breathing rate and depth whether or not cardiac vagal tone changed, and respiration was left uncontrolled in most studies pooled by Zahn et al. (2016) and Wesarg et al. (2022).
• Treating RSA/HRV as ‘vagal tone’ is contested: Grossman (2023) calls RSA = cardiac vagal tone, and still more RSA = general vagal tone, a category mistake, since cardiac, pulmonary and gut vagal activity do not covary.
• Individual-difference prediction is weak in the wiki’s meta-analyses: resting HRV and self-control r = 0.15, falling to a nonsignificant r = 0.04 after publication-bias correction (Zahn 2016); no overall link between childhood adversity and baseline or reactive vmHRV (Wesarg 2022); RSA withdrawal unrelated to children’s social functioning (Graziano & Derefinko 2013).
• Metric and computation choices are heterogeneous (HF vs RMSSD vs SDNN; reactivity as task-mean vs change score), and some results depend on them, e.g. Wesarg’s reactivity effect appears only for task-mean-level scores.
• Linear time/frequency-domain metrics cover vagal influence only; sympathetic contribution and nonlinear inter-beat-interval complexity are not captured.

Connections

• Key studies: Zahn et al. (2016), Wesarg et al. (2022), Graziano & Derefinko (2013), Grossman (2023), Porges (2023), Kerdabadi et al. (2024), Banellis et al. (2026), Greenwood & Garfinkel (2025), Bonaz et al. (2021), Weng et al. (2021)
• Related methods: device-guided breathing (DGB), Heartbeat-evoked potential, Heartbeat detection task, Skin conductance response (SCR), Electrogastrography (EGG), Multidimensional Experience Sampling (MDES)

HRV is the wiki’s main efferent cardiac measure: it reads how the autonomic nervous system, chiefly the vagus, modulates the heart, and asks nothing of the participant. That separates it from the heartbeat-perception measures (Heartbeat detection task, Heart Rate Discrimination Task (HRDT)) and from cortical processing of the heartbeat (Heartbeat-evoked potential). The construct its vagal components are taken to index, and the dispute over what that index may mean, are kept on the concept page Respiratory sinus arrhythmia (RSA).

How the wiki’s sources use it

• Trait predictor of regulation. Zahn et al. (2016) test the Neurovisceral Integration claim that resting HRV reflects prefrontal inhibitory capacity. The pooled effect on laboratory self-control is small (r = 0.15) and nonsignificant after bias correction (r = 0.04); HRV metric (HF, RMSSD, SDNN) did not moderate it.
• Developmental marker. Wesarg et al. (2022) find no overall association between childhood adversity and baseline or reactive vmHRV, with small effects only in clinical samples and for maltreatment (see Early adverse life events). Graziano & Derefinko (2013) link RSA withdrawal to fewer externalizing, internalizing and academic problems, but not to social functioning.
• State correlate of thought content. Banellis et al. (2026) record PPG during resting fMRI: body-focused thought (Body-wandering) goes with higher heart rate and lower RMSSD, cognitive content with higher RMSSD.
• Clinical profile. Greenwood & Garfinkel (2025) list reduced resting HRV in anxiety, PTSD, schizophrenia and EUPD (see Interoceptive psychopathology); Bonaz et al. (2021) include HRV among candidate interoceptive biomarkers, while conceding few have found broad clinical use.
• Intervention outcome and mechanism. Weng et al. (2021) report RSA improving after MABT. Kerdabadi et al. (2024) pool “polyvagal” interventions, including HRV biofeedback and device-guided breathing (DGB), that mostly act through generic parasympathetic/HRV mechanisms.

The RSA-as-vagal-tone caveat

Polyvagal Theory rests on RSA as a readout of the ventral vagal brake (Porges 2023). Grossman (2023) argues that RSA only approximates cardiac vagal tone, and only with respiration controlled; that it says nothing about vagal outflow to lungs or gut; and that emotion effects on RSA are largely mediated by breathing. The Porges-Bohrer metric is itself respiration-dependent (r of about 0.84 with respiratory rate). The practical rule: an HRV change licenses “cardiac vagal influence probably changed”, not “vagal tone” or “autonomic state” changed. See Is Polyvagal Theory scientifically valid?.