On choosing heart-rate-variability (HRV) metrics to reflect vagus-nerve functioning: The construct counts, not the...

Diverse strategies exist in psychophysiology to quantify respiratory heart-rate variability (RespHRV), previously termed respiratory sinus arrhythmia, including high-frequency spectral HRV (HF-HRV), the Porges-Bohrer moving polynomial filter approach (MPF_P-B), and peak-valley RespHRV (RespHRV_PV). RMSSD is also more recently treated as RespHRV-relevant, although it is more appropriately a broad HRV descriptor. Given the proliferation of numerous quantification methods, the field has reached a decision point about which metric might be "superior." Divergent evidence has further fueled this debate and has led to confusion, including claims that some indices are preferable because they show larger blockade effects, might function as respiration-free measures, or inherently handle nonstationarity and related statistical properties better. This perspective article argues that the main bottleneck is not the RespHRV metric itself at the algorithm level. Instead, the core issue is construct validity and the validity of inferences drawn from RespHRV metrics, which must be anchored in the physiological origin of RespHRV as cardiorespiratory coupling. When comparable preprocessing pipelines are used, commonly used RespHRV metrics tend to be highly correlated, and apparent divergences often trace back to non-uniform processing, design mismatches, or over-interpretation of limited evidence. We review why vagal blockade findings do not necessarily establish a purer vagal tone metric, especially when blockade is partial, dosing is not titrated, and sympathetic influences are unconstrained. We also emphasize that respiration cannot be treated as optional or as nuisance variance: RespHRV is inherently respiratory, and claims of respiration-free metrics typically rely on underpowered, between-individual analyses or null results. Finally, we clarify that nonstationarity should be evaluated in terms of estimation bias within physiologically plausible ranges, and that detrending or filtering can, itself, introduce bias. Overall, the field's priority should shift from metric preference debates to careful measurement of respiration, transparent and uniform preprocessing, appropriate method-comparison statistics, and cautious interpretation at the construct level.