Do Breathing Exercises Actually Increase HRV?

This article covers what controlled research has found about slow, paced, or resonance-frequency breathing and its effect on HRV metrics. It does not cover general relaxation apps, wearable-guided breathing features as products, or long-term cardiovascular outcomes.

Research on paced breathing at a person's resonance frequency, the pace where breathing, heart rate, and blood pressure oscillations line up most efficiently, has repeatedly shown it raises HRV and increases baroreflex sensitivity during and after practice. This has been demonstrated across separate controlled trials using different ways of finding that pace. The effect is measured in specific HRV parameters like SDNN and baroreflex sensitivity, not as a universal experience, and not everyone engages with the training the same way.

Why this keeps coming up

Someone starts a slow-breathing routine because a wearable told them their HRV number is low, or because they read that a few minutes of paced breathing a day is supposed to move the needle. Then the number on the app doesn't budge for a week, or it does, and they wonder if that's the breathing or just noise. The gap between "breathing exercises raise HRV" as a stated fact and what actually happens on a given Tuesday morning is where most of the confusion lives.

Part of the confusion is that HRV itself is a stack of different metrics measured in different ways, and the number itself means less than people assume. Part of it is that resonance-frequency breathing, the specific technique studied here, is not the same as just breathing slowly, and the two get conflated constantly.

What "resonance frequency" actually means here

The studies behind this thesis aren't testing "slow breathing" as a vague category. They're testing a specific pace, roughly the frequency at which the body's baroreflex loop (the system linking blood pressure and heart rate) naturally oscillates, somewhere near 0.1 Hz for many adults but not identical across people. One trial identified this pace by locating the peak of the low-frequency band directly and compared it against a more standard method of estimating resonance frequency, finding the baroreflex sensitivity gain was larger with the direct spectral approach [6]. That detail matters because it suggests the exact method used to find someone's resonance pace isn't interchangeable, even within a single study population.

A separate four-week training study in young adults found measurable gains in SDNN and total power, alongside lower perceived stress and improved performance on a cognitive task, when resonance breathing was practiced over that stretch rather than as a single session [7]. That's a different claim than "HRV goes up while you're doing it," and it's worth holding the two apart.

Then there's a wrinkle in the assumed mechanism itself. For years the working model was that during resonance breathing, heart rate and breathing move exactly out of phase with each other in a clean 180-degree relationship. When researchers actually measured that phase relationship across a range of adult ages, they found something closer to 109 degrees, with heart rate changes actually leading the breath rather than trailing it in the assumed pattern. The oscillations still happened, the baroreflex was still stimulated, but the underlying timing wasn't what practitioners had been describing [8]. That's a good reminder that a technique can reliably produce an effect while the textbook explanation of why still gets revised.

Why some people don't see it move

If paced breathing at resonance frequency has this level of support, why does it sometimes feel like nothing is happening on a given wearable readout? One narrative review proposes a framework it calls autonomic non-responsiveness, essentially naming the situation where the expected physiological engagement during HRV biofeedback training is weakened, absent, or doesn't translate into a measurable benefit. It points to contributors like reduced autonomic flexibility, impaired baroreflex function, existing disease burden, fatigue, and dysfunctional breathing patterns as reasons the same technique doesn't land the same way for everyone [0]. That's a conceptual framing rather than a new trial, but it directly addresses the mismatch between "this technique is well-studied" and "it isn't working for me right now."

There's also the separate question of whether breathing-based HRV training holds up against other stress-reduction approaches. A randomized trial comparing physical activity, mindfulness meditation, and HRV biofeedback via slow breathing over five weeks of home practice found an overall beneficial effect across all three arms on stress, anxiety, depressive symptoms, wellbeing, and sleep quality, without singling out one method as clearly superior [1]. That's a useful check against assuming breathing-based HRV training is uniquely powerful rather than one effective option among several. For a closer look at whether the anxiety-reduction piece specifically holds up, the evidence on HRV biofeedback and anxiety is worth a separate look, and the feasibility question for people with limited physical function is addressed in a small pilot trial in adults with tetraplegia, which found the training arm spent more time in the target low-frequency range and reported greater anxiety improvement than the monitoring-only control, with full protocol completion and no withdrawals [4].

What this doesn't tell you about your own number

None of this evidence says a rising HRV reading during or after breathing practice is a stand-in for feeling less stressed, or that a flat reading means the technique failed. HRV is shaped by measurement method, time domain versus frequency domain choice, and even the device doing the recording, and how wearables actually capture HRV differs enough between sensor types that comparing a resonance-breathing session's numbers across different devices can be misleading on its own. Guidance for how HRV should be measured and reported in research settings underscores just how many methodological choices sit behind a single number, which is part of why single-session shifts on a consumer wearable don't map cleanly onto the controlled lab conditions these studies used [11].

It's also worth separating short-term, in-session HRV changes from the broader question of resting HRV as a marker of autonomic balance and cardiovascular risk over time, which is a different body of evidence entirely, tied more to longer-term influences like exercise frequency than to a single breathing session [10].

Common questions