HRV Biofeedback: Evidence and Limits

What it is

Heart rate variability (HRV) is the small variation in time between heartbeats. HRV biofeedback is a training method in which a person sees a live display of heart signals, usually through a sensor, and uses slow, paced breathing to influence it. It is sometimes called resonance breathing or resonance-frequency breathing. It is different from the single overnight HRV number many wearables report: that number is an estimate from a sensor, while biofeedback is a guided practice done with real-time feedback. This page does not give breathing settings or instructions. For the practice itself and its safety, see the Therapies page on box breathing, resonance breathing and HRV biofeedback.

Evidence labels used: promising but limited, mixed and insufficient evidence.

Stress and anxiety

A 2017 meta-analysis of 24 studies (484 people) found a pre-post effect size of 0.81 and a between-group effect size of 0.83 versus control conditions, which the authors described as a large reduction in self-reported stress and anxiety. They added that more well-controlled studies were needed. [1] Label: promising but limited. Self-reported symptoms are affected by expectations, and many studies were small.

Depression, stress and remote programs

A 2025 systematic review and meta-analysis of 18 studies (1,352 people) looked at remote HRV biofeedback. It found a medium effect on depression (g = -0.41) and on HRV (g = 0.443) compared with controls, with large variation between studies. The effect for stress was not significant. [2] Label: mixed. The two reviews disagree on stress, and the reviews pooled different studies, delivery methods and comparison groups, so they are not directly comparable.

PTSD in military service members

A 2024 meta-analysis of five studies (95 service members) found a moderate to large mean reduction in PTSD symptoms (Hedges g = -0.557). [3] Label: insufficient evidence to draw firm conclusions, because the number of studies and people is small. Biofeedback is not a substitute for trauma care from a clinician.

What the research does not show

• That a higher HRV number is better for everyone. There is no single ideal HRV, and wearable HRV numbers have no universal clinical meaning. See HRV trends for recovery and HRV score vs subjective recovery.
• That a consumer app or device is as effective as the programs studied. Reviews pool many different products and protocols, and we did not find evidence on any specific consumer device to cite. Proper evidence not found for device-specific trials in the reviews we read.
• Accuracy is not benefit. A device may measure heart signals well and still not improve your health.

How to evaluate a claim

Check whether the trial had a control group, whether results were self-reported, how many people took part and whether the product being sold was the one studied. See how to evaluate a biohack and how to evaluate a health device claim.