Why your HRV response to vagus nerve stimulation isn’t always the same?

Transcutaneous auricular vagus nerve stimulation (taVNS) is attracting growing attention—both in research and among everyday users—because it offers a non invasive way to engage vagal pathways involved in stress regulation, recovery, and autonomic balance.

At ZenoWell, we’re seeing this momentum clearly. Our taVNS product is being used by more and more people to support heart rate variability (HRV). In our earlier reproducibility work using our research-grade taVNS device (BrainCLOS) in a university-student cohort, we observed that in 60 participants, our taVNS research-grade setup increased HRV by about 12% compared to sham, using RMSSD as the primary metric (Xiao et al., 2025, OHBM Annual Meeting, Brisbane, Australia)

At the same time, many direct users tell us something important: their HRV outcomes are not always consistent. Some people see clear improvement, some see minimal change, and some even see HRV go in the “wrong” direction on certain days.

In this blog, we want to share how we think about these inconsistencies—based on the latest science—so users can interpret their data more accurately and so we can move toward a more personalized, reliable taVNS experience.

Baseline physiology matters more than most people think

A major theme emerging from recent taVNS research is that “taVNS vs sham” effects can look small or inconsistent at the group level, but become much clearer when you consider where a person starts from.

1. Low baseline HRV (low RMSSD) may mean “more room to improve”. In a 2025 randomized, single-blind crossover trial in Major Depressive Disorder (MDD) and healthy controls, Schiweck et al. did not find strong taVNS effects when comparing MDD vs controls directly. However, when they performed a post-hoc analysis stratifying participants by baseline RMSSD (low vs high), they observed notably different patterns:

• Low RMSSD group: taVNS tended to restore a blunted cardiac stress response, suggesting improved physiological responsiveness or flexibility under stress.

• High RMSSD group: unexpectedly, taVNS showed the opposite direction—heart rate and inflammatory markers (TNF α) increased and vagally mediated HRV decreased under taVNS compared to sham.

This is a key message for real-world users: the same stimulation can move physiology in different directions depending on baseline vagal tone.

2. High sympathetic activation may also predict sensitivity. A separate paper published in 2026 by Percin et al. focuses directly on whether baseline autonomic state affects taVNS outcomes. The overall conclusion is consistent with the idea above: individuals with lower parasympathetic activity and/or higher sympathetic activation may be more “responsive” to taVNS, with clearer shifts in autonomic indices after stimulation.

Taken together, these studies support a practical working hypothesis: If your baseline HRV is lower (or your physiology is more sympathetically “activated”), you may be more likely to see measurable changes from taVNS. If your baseline RMSSD is already high, your response may be smaller, different, or more context-dependent.

Importantly, this does not mean taVNS is “only for low HRV.” It means that baseline biology helps determine what outcome you should expect and how to interpret day-to-day variability.

Stimulation parameters can determine what HRV metric changes (and whether RMSSD changes at all)

Another reason results differ is simple: “taVNS” is not one single intervention. Frequency, pulse width, intensity, timing, and stimulation placement can all change the physiological response.

A 2025 randomized crossover controlled trial by Atanackov et al., in healthy adults systematically tested six active protocols (10 Hz vs 25 Hz, combined with 100/250/500 μs pulse width) and sham. The study found that certain parameter combinations significantly increased SDNN (a measure of overall HRV), but none significantly changed RMSSD.

This is highly relevant to real-world tracking because many people focus on RMSSD only. Depending on your protocol and context, you might see:

• shifts in overall variability (SDNN) without a clear RMSSD change, or

• changes that appear during specific time windows (e.g., later in stimulation or during recovery), or

• effects that become detectable primarily in people with certain baseline profiles (low RMSSD, high sympathetic activity, etc.).

So when users say “taVNS didn’t improve my HRV,” the more accurate question is: Which HRV metric are you looking at, measured under what conditions, with what stimulation parameters, and relative to what baseline?

More interestingly, Stimulation parameters & Sex dependent?

Recent preclinical work by Barbetti et al., (2025) suggests that biological sex may be an underappreciated driver of variability in taVNS-to-HRV outcomes. In adult rats, taVNS produced sex- and frequency-dependent effects (HR and vagally mediated HRV increased most clearly at 6 Hz in males and at 20 Hz in females), and notably, vmHRV increases were observed primarily with right-sided auricular stimulation. 

Together, these findings imply that “best” taVNS parameters—and even which ear to stimulate—may need to be personalized rather than assumed to generalize across users.

HRV measurement in daily life — why wearables can be confusing (and why we recommend trend tracking)

We also want to address a major source of “inconsistent results” that isn’t actually about taVNS at all: HRV measurement.

HRV is extremely sensitive. It can change substantially due to:

• posture (standing vs sitting)

• breathing pattern

• emotional state (calm vs stressed)

• timing relative to meals (fasted vs fed)

• sleep quality and sleep debt

• caffeine, alcohol, hydration

• time of day and acute stressors etc.

That sensitivity is what makes HRV useful—but also what makes single readings easy to misinterpret.

Research-grade ECG vs wearable HRV In research studies, HRV is typically derived from ECG/EKG recorded at high sampling rates (often 500–1000 Hz or higher) with standardized time windows (commonly 5 minutes) and artifact correction. For example, the depression trial used ECG sampled at 1024 Hz and analyzed structured baseline/stress/recovery segments.

Wearables are different. They are great for convenience and long-term tracking, but they often estimate HRV from optical PPG and may compute values from short, opportunistic windows under uncontrolled real-world conditions. That makes “one data point” much noisier than a controlled ECG-derived 5 minute segment.

What this means practically Because sampling and measurement windows differ across devices—and because HRV is inherently context-sensitive—comparing a single wearable HRV value to a research-grade ECG RMSSD is often not an apples-to-apples comparison.

Our recommendation: track trends, not single values For most users, we recommend evaluating taVNS impact by looking at longer-term trends: Track HRV over weeks rather than days
• Compare similar contexts (same time of day, similar posture, similar pre-/post-meal state)
• Focus on trend direction and stability rather than one “good” or “bad” point

This approach dramatically reduces false conclusions caused by posture, mood, meals, and wearable measurement noise.

Where ZenoWell is going — personalized parameters, clear targets, and closed-loop optimization

The combined message from these studies is not “taVNS doesn’t work.” It’s that taVNS is not one-size-fits-all, and “the right settings for the right person” matters.

Our long-term goal at ZenoWell is to help users move from: generic settings and ambiguous outcomes
to
• personalized stimulation parameters, paired with
• clear HRV improvement targets, and ultimately
• a closed-loop feedback system that can optimize parameters based on your physiology over time.

Finally, HRV is not the whole mechanism. Many people use taVNS to support sleep, stress management, pain, and fatigue. Autonomic regulation and HRV are important, but they are not the only pathways involved—taVNS also engages brainstem-cortex networks and downstream systems that can influence inflammation, neuromodulation, attention, and emotional regulation in ways that HRV alone may not fully capture.

Stay tuned—we’ll share more about these mechanisms and how we’re translating them into a more personalized ZenoWell experience.

References:

1. Atanackov, P., Peterlin, J., Derlink, M., Kovačič, U., Kejžar, N., & Bajrović, F. F. (2025). The Acute Effects of Varying Frequency and Pulse Width of Transcutaneous Auricular Vagus Nerve Stimulation on Heart Rate Variability in Healthy Adults: A Randomized Crossover Controlled Trial. Biomedicines, 13(3), 700.

2. Barbetti, M., Ottaviani, C., Thayer, J. F., Sgoifo, A., & Carnevali, L. (2025). Sex differences in heart rate and heart rate variability responses to transcutaneous auricular vagal nerve stimulation in rats. Autonomic Neuroscience, 257, 103237.

3. Percin, A., Ozden, A. V., Yenisehir, S., Pehlivanoglu, B. E., & Yılmaz, R. C. (2026). Does Baseline Autonomic Nervous System Activity Affect the Outcomes of Transcutaneous Auricular Vagus Nerve Stimulation?. Archives of Medical Research, 57(3), 103324.

4. Schiweck, C., Aichholzer, M., Brandt, E., Schneider, M., Meyer, K., Hamzehloiya, T., ... & Edwin Thanarajah, S. (2025). The heart knows best: baseline heart rate variability as guide to transcutaneous auricular vagus nerve stimulation in depression. Translational Psychiatry.