How to Improve Your HRV: What the Evidence Supports (and the Noise to Ignore)

What actually raises HRV?

Heart rate variability is the variation in the time between consecutive heartbeats, and the vagally mediated part of it — the part your app usually reports as RMSSD — reflects parasympathetic input to the heart. If you have not met the metric before, start with what HRV actually measures; this article is about what changes it.

Ranked by the quality of the evidence behind them, rather than by how often they are posted about:

Does exercise improve HRV?

Yes, and it is the best-evidenced lever you have. Sandercock and colleagues pooled the training studies available at the time and found that exercise training raised high-frequency power — the vagally mediated band — with an effect size of d = 0.48 across 13 studies and 322 cases, and lengthened the RR interval (that is, lowered resting heart rate) with d = 0.75 across 12 studies and 298 cases. Both were statistically robust.

Two caveats that the same analysis makes explicit. First, the response was smaller in older participants for the RR interval, with a similar trend for HF power. Second, the relationship between the change in HF power and the change in resting heart rate was weak, which means training bradycardia is not purely a vagal story — so do not expect your HRV and your resting heart rate to move in perfect lockstep.

The practical reading: consistent aerobic volume at an intensity you recover from is what raises HRV over months. Piling on hard sessions does the opposite in the short term — which is the whole basis for using HRV to time hard days at all.

Does slow breathing raise HRV?

It does, more reliably than almost any other acute intervention. Laborde and colleagues screened 1,842 abstracts and included 223 studies of voluntary slow breathing, split across measurements taken during the breathing session (172 studies), immediately after a single session (16) and after a multi-session intervention (49). Vagally mediated HRV increased in all three contexts, with few adverse effects expected — the authors' conclusion was that slow breathing is a reasonable low-tech, low-cost prevention technique.

The rate that matters is roughly six breaths a minute. At about that pace, the oscillations in breathing, blood pressure and heart rate line up at around 0.1 Hz and reinforce each other; Lehrer and Gevirtz's review identifies strengthening of the baroreflex as the best-supported mechanism behind heart rate variability biofeedback, with an effect on the vagal afferent pathway as a further proposal. In practice that is about a five-second inhale and a five-second exhale, done calmly and without straining.

Formal HRV biofeedback — breathing at your own resonance frequency with live feedback, usually weekly for several weeks — has been tested mostly on symptoms rather than on HRV itself. Goessl, Curtiss and Hofmann's meta-analysis of 24 studies and 484 participants found a large reduction in self-reported stress and anxiety (Hedges' g = 0.81 within-group, g = 0.83 against control), not moderated by the number of sessions or by whether participants had an anxiety disorder. That is a real finding about symptoms; it is not the same claim as "biofeedback permanently raises your resting HRV", and the honest summary is that the symptom evidence is stronger than the trait-HRV evidence.

How much does alcohol lower your HRV?

More than most people expect, and at doses most people consider trivial. Pietilä and colleagues analysed beat-to-beat recordings from 4,098 Finnish employees in ordinary life, comparing each person's nights with and without alcohol. Alcohol was associated with increased sympathetic and decreased parasympathetic regulation during the first three hours of sleep in a dose-dependent way. Their HRV-derived recovery measure fell on average by 9.3 percentage points after a low dose (up to 0.25 g/kg), 24.0 after a moderate dose (0.25–0.75 g/kg) and 39.2 after a high dose.

Two details deserve attention. The effect was present even at the low dose — for an 80 kg adult, 0.25 g/kg is roughly two standard drinks. And being physically active did not protect against it; if anything the suppression was stronger in younger participants. If you want a single, immediate, visible change in your HRV data, an alcohol-free stretch of a week or two is the intervention most likely to produce one.

What does sleep loss do to HRV?

It shifts the autonomic balance in the direction you would expect, though less uniformly than the wellness framing suggests. Zhang and colleagues pooled 11 randomised controlled trials of sleep deprivation covering 549 participants and found a significant fall in RMSSD — the vagal index most consumer apps report — alongside significant rises in LF power and the LF/HF ratio. SDNN fell and HF fell, but neither reached statistical significance in the pooled analysis.

So: the measure your ring or watch shows you is the one that reliably drops after a bad night. That is a genuine reason to treat sleep as an HRV intervention rather than as an outcome, and it is why the practical advice in how to improve sleep quality does double duty here.

What can't you change? Age, sex and your own baseline

A large part of your HRV number was decided before you did anything. Umetani and colleagues measured 24-hour HRV in 260 healthy people aged 10 to 99 and found a clear age decline whose shape depends on the metric: using RMSSD and pNN50, HRV dropped fastest, reaching 47% and 24% of second-decade values by the sixth decade and then levelling off, while SDNN declined gradually to about 60% of baseline by the tenth decade. Below the age of 30, HRV was lower in women than in men; the difference shrank after 30 and had disappeared by 50.

Between-person spread is larger still. Nunan and colleagues' systematic review of 44 studies and 21,438 participants found published short-term HRV values sitting below the 1996 Task Force norms, with interindividual variation running as high as 260,000% for spectral measures. Shaffer and Ginsberg make the corresponding methodological point: 24-hour, short-term and ultra-short-term norms are not interchangeable, and baseline values depend on recording length, age and sex.

The conclusion is not "HRV is useless". It is that comparing your number to someone else's, or to a population chart, tells you almost nothing. Comparing this week's average to your own last month tells you a great deal.

How to measure HRV so an improvement is visible

Most people fail to see their HRV improve because of measurement noise, not because the intervention failed. Plews and colleagues, working with Olympic and world-champion endurance athletes, make the case for appropriate averaging techniques — rolling weekly averages rather than single daily values — and for watching the spread of the daily readings as well as the mean. They also document HRV saturation in very fit athletes, where HRV can fall while resting heart rate also falls, so higher is not unconditionally better at the top end.

A protocol that works:

1. Same time, same posture, every day. On waking, before coffee, before looking at your phone. Position changes HRV substantially, so pick lying or seated and stay with it.
2. Use one metric. RMSSD or its logarithm is the usual choice for short recordings because it is the least affected by breathing rate.
3. Judge on a 7-day rolling average. Compare this week to last month, never today to yesterday.
4. Log the obvious confounders — alcohol, late sessions, illness, travel — or the average will move for reasons you cannot reconstruct later.

Can HRV tell you when to train hard?

There is one well-designed trial on exactly this. Vesterinen and colleagues split 40 recreational endurance runners into a group following a predefined programme and a group whose hard sessions were scheduled only when morning HRV sat within their individual normal range. Over eight weeks the HRV-guided group did significantly fewer hard sessions (13.2 ± 6.0 versus 17.7 ± 2.5), and their 3,000 m time improved by 2.1% (p = 0.004) while the predefined group's 1.1% improvement did not reach significance. VO2max improved in both groups.

Read honestly, that is a promising result about timing hard sessions, not proof that HRV-guided training beats a good plan: the between-group difference in the 3,000 m was small, and VO2max rose in both arms. The useful takeaway is that skipping a hard session on a genuinely suppressed morning cost these runners nothing.

What to track alongside HRV

HRV on its own is a number without a cause. Every one of the interventions above shows up in HRV only through something you can also log directly: a drink, a short night, a hard session, an illness. Track those next to the reading and the number becomes explainable rather than mysterious.

• Alcohol, by day and amount — the single clearest signal you will find in your own data, and the fastest to verify.
• Sleep duration and timing, not just a sleep score — RMSSD is the measure that falls after deprivation, so it should track your worst nights.
• Training load and session intensity, so a suppressed morning can be attributed to yesterday's intervals instead of being treated as a mystery.

tr8ck's sleep module pulls HRV and resting heart rate from Apple Health or Health Connect and puts them on one screen with what you logged the day before — alcohol, training, mood, food — so the weekly average has context attached. You can also log it all without a wearable, which is what tracking sleep quality manually covers. The point is not the number; it is which of your own habits moves it.

FAQ

Medical disclaimer: This article is for informational purposes only and does not constitute medical advice. Heart rate variability is not a diagnostic test, and a persistently low or falling HRV can have medical causes. If you have a heart condition, an irregular heartbeat, take medication that affects heart rate, or your HRV falls alongside symptoms such as breathlessness, chest pain or prolonged fatigue, speak to a qualified healthcare professional rather than adjusting training on your own.