6 Pitfalls of Interpretation: Confounders
When you get the result “HRV went down”, it does not necessarily mean a change in the autonomic nerves (the vagus). This chapter organizes the main confounders that affect HRV.
6.1 1. Heart rate itself
HRV and heart rate are strongly correlated (the higher the heart rate, the smaller the HRV). There are two reasons for this.
6.1.1 Physiological reason
It is mainly the same vagus nerve that both lowers the heart rate and creates the beat-to-beat fluctuation. When vagal activity is high, heart rate is low and HRV is large. This is less a “confounder” than two manifestations of the same cause.
6.1.2 Mathematical reason
RR interval and heart rate are reciprocally related.
\mathrm{RR}\,[\mathrm{ms}] = \frac{60000}{\mathrm{HR}\,[\mathrm{bpm}]} \quad\Rightarrow\quad \Delta \mathrm{RR} \approx -\frac{60000}{\mathrm{HR}^2}\,\Delta \mathrm{HR}
For the same ±3 bpm fluctuation in heart rate,
- at a heart rate of 60 bpm (RR 1000 ms): the RR fluctuation is about ±50 ms
- at a heart rate of 90 bpm (RR about 667 ms): the RR fluctuation is about ±22 ms
so the RR-interval fluctuation looks smaller just because the heart rate is higher. For this reason, methods that correct HRV for heart rate (or mean RR) have also been proposed (e.g., the coefficient of variation SDNN / mean NN, Sacha’s correction).
When comparing HRV between two groups or conditions, report heart rate as well, and check whether the difference can be explained by the difference in heart rate.
6.2 2. Respiration
As we saw in Chapter 3, HF power depends strongly on respiratory rate and tidal volume.
| Change in breathing | Effect on HF | Example of misinterpretation |
|---|---|---|
| Breathing slows (< 9 breaths/min) | The respiratory fluctuation moves into the LF band → HF↓, LF↑, LF/HF↑↑ | “Sympathetic dominance” |
| Breathing speeds up (> 24 breaths/min) | The respiratory fluctuation exceeds the upper HF limit (0.4 Hz) → HF↓ | “Vagal activity decreased” |
| Deep breathing | RSA becomes larger → HF↑ | “Vagal activity increased” |
| Talking, yawning, sighing | Irregular large fluctuations | — |
In studies of meditation or breathing techniques, measurements during exercise, tasks involving speech, and so on, the effect of respiration must always be considered. Countermeasures include recording respiratory rate, controlling breathing (metronome breathing), and shifting the HF band according to respiratory rate.
6.3 3. Posture
When going from supine → sitting → standing, the baroreflex responds to reduced venous return, resulting in
- vagal activity↓, sympathetic activity↑
- heart rate↑, RMSSD and HF↓
Only values measured in the same posture can be compared with each other. There are also methods that use the response to standing itself for assessment (orthostatic tests).
6.4 4. Age, sex, fitness, genetics
- Age: HRV declines greatly with age, especially from the 20s to the 50s
- Sex: young women tend to have a higher proportion of the HF (vagal) component than men
- Aerobic fitness: endurance-trained people have a low resting heart rate and large HRV
- Genetics: a substantial part of inter-individual differences in HRV is explained by genetic factors (twin studies)
6.5 5. Time of day, sleep, lifestyle
- Circadian rhythm: vagal activity increases at night (especially during non-REM sleep), and HRV rises
- Sleep stage: stably high during non-REM sleep and unstable during REM sleep
- Meals: after eating, heart rate rises and HRV falls
- Caffeine, alcohol, smoking: alcohol in particular greatly lowers nighttime HRV
- Physical condition: infection/fever, dehydration, sleep deprivation, menstrual cycle
6.6 6. Drugs
- β-blockers: heart rate↓, HRV↑ (toward vagal dominance)
- Drugs with anticholinergic effects (tricyclic antidepressants, some antihistamines, etc.): HRV↓
- Calcium channel blockers, antiarrhythmics, and others also have effects
6.7 7. Measurement and analysis conditions
- Measuring device (ECG or PPG), sampling frequency
- Method and rate of correcting ectopic beats and artifacts
- Recording length, choice of analysis segment
- Spectral estimation method (FFT / AR / Lomb–Scargle), resampling frequency, detrending
6.8 Example of standard measurement conditions
Commonly used conditions for measuring resting HRV in research or daily monitoring include:
- Fix the time of day (e.g., right after waking, after urinating)
- Fix the posture (supine or sitting), with a few minutes of rest before measuring
- Fix the recording length (e.g., 5 min; for wearables, 1–5 min or during nighttime sleep)
- Avoid caffeine, meals, and exercise before measuring
- If possible, record respiratory rate, or breathe at a constant pace
- Avoid talking and moving
When HRV changes, check first:
- Has the heart rate changed?
- Has breathing (rate, depth) changed?
- Are posture, time of day, and recording length the same?
- Is the correction of ectopic beats and artifacts appropriate?
Only changes that cannot be explained by these should be interpreted as autonomic changes.