How to improve your HRV: what research links to it, and how to read your own trend

Can you really improve your HRV, and by how much?

Partly, and for most habits the honest answer is “a little, and it depends on you”. This guide is about action: which habits studies have linked to a higher or lower heart rate variability (HRV), how long anything took, and how to read your own numbers. For what HRV is, and how it varies with age, start with the heart rate variability guide. Those definitions are not repeated here.

The link with the most quantitative support in the sources read is exercise. A meta-analysis from the UK pooled 13 studies and 322 cases and found that exercise training raised the high frequency component of HRV, with an effect size of 0.48 and a confidence interval of 0.26 to 0.70.⁸ We describe an effect size of 0.48 as modest, and the pooled studies varied in who took part and how they trained.

Two more cautions. First, a higher HRV is not a prize that everyone should chase. The sources read for this guide give no target number, and the sensible comparison is with your own earlier readings. Second, nothing here is a promise. The studies describe groups of people over a set period, and your result may differ.

What the research links to HRV: habit by habit

The table sets out what each kind of study reported. It is a list of links, not a ranking, and no weighting has been added.

Alcohol: the clearest short term link

In 4,098 Finnish employees followed in daily life, alcohol was associated with weaker parasympathetic regulation in the first 3 hours of sleep, and the effect grew with the dose.¹³ A recovery measure derived from HRV fell on average by 9.3, 24.0 and 39.2 percentage units after low, moderate and high intakes.¹³ Regular exercise and young age did not protect against this.¹³

The study defines the doses in grams of pure alcohol per kilogram of body weight: up to 0.25 for low, over 0.25 to 0.75 for moderate and over 0.75 for high.¹³ For someone weighing 70 kg (about 11 stone), that is 0.25 × 70 = 17.5 g at the low boundary and 0.75 × 70 = 52.5 g at the high one. The NHS counts 8 g of pure alcohol as one unit, so those are about 2.19 and 6.56 units (17.5 ÷ 8 and 52.5 ÷ 8).¹ The 70 kg body weight is only an illustration.

To count units, the NHS formula is strength (ABV) multiplied by volume in millilitres, divided by 1,000. A pint (568 ml) of 5.2% beer works out at 5.2 × 568 ÷ 1,000 = 2.95 units.¹ The NHS advises against drinking more than 14 units a week on a regular basis, and spreading it over 3 days or more if you drink that much.¹

A review of 33 articles found that acute alcohol reduced resting HRV in healthy people.¹⁴ The same review reported that, in people with alcohol use disorder, a clinical group and not the general public, resting HRV improved only after at least 4 months of abstinence.¹⁴ That 4 months does not apply to an ordinary drinker.

The NHS adds that alcohol can make you sleepy at first but lead to poor sleep later in the night.³ It also advises not smoking or drinking alcohol, tea or coffee for at least 6 hours before bed.³ For more on this, see how to improve sleep quality.

Exercise: helpful, but the type and dose matter

Endurance training normally raises HRV, according to a 2024 narrative review, while intensive training and runs of competitions can lower it. The same review says strength training usually changes nothing in healthy people.⁶ Straight after a session, HRV dips first.⁶

One small randomised trial of 20 inactive adults found that a 12 week programme of high intensity interval training could raise short term HRV (SDNN).⁹ The abstract also reports figures for the gap between the two groups that do not add up, so this guide does not quote that gap. Treat the finding as an indication only, given the size of the trial.

A systematic review of 13 studies in healthy adults aged 60 and over found a favourable effect on the heart’s autonomic control from endurance, coordinative and multimodal training, but not from resistance training.¹⁰

A meta-analysis in people with type 2 diabetes found that the duration and frequency of training did not change the HRV benefit.¹¹ That is a group with a medical condition. It should not be read as meaning that the amount of exercise does not matter for everyone.

For activity levels, the NHS recommends adults aged 19 to 64 do at least 150 minutes of moderate intensity activity a week, or 75 minutes of vigorous activity.² Spread over 5 days, 150 minutes is 30 minutes a day.² The NHS says to speak to your GP first if you have not exercised for some time or have a medical condition.²

Sleep, stress and breathing

A meta-analysis of 11 randomised trials with 549 participants found RMSSD fell significantly after sleep deprivation, while the fall in SDNN was not significant.¹⁵ That is deprivation under study conditions. It does not describe an ordinary short night, and the same review of factors notes that a lower HRV with sleep disorders is not supported by the literature.⁶ The NHS says most healthy adults need around 7 to 9 hours of sleep a night, and advises going to bed and getting up at similar times each day.³ For the practical side, see how to improve sleep quality and how much sleep you need.

Stress is linked with less parasympathetic activity, but the review read gives no figure and notes that biofeedback effects were seen only in the short term.⁶ On breathing, a meta-analysis of 223 studies found that voluntary slow breathing raised vagal HRV during the session, just after it, and after a course of sessions.¹² The abstract does not give a size or a timescale, so none is given here.

How long before a change shows?

Only two durations were found in the sources read, and neither is a general rule for healthy adults.

The practical point: expect to judge a change over weeks, with the same method, and not over a few nights. Whether you will see a change at all, and when, cannot be predicted from these studies.

Why a night can be low: what the sources do and do not say

People often ask what lowers HRV. From the sources read, the supported answers are acute alcohol ¹³¹⁴, experimental sleep deprivation ¹⁵, mental or work stress ⁶, night shift work ⁶ and very intensive training ⁶. The NHS also lists lack of sleep, stress and anxiety, alcohol, caffeine and intense exercise among common triggers of palpitations, which is a different thing from HRV.⁴

Nothing read for this guide shows how HRV behaves in illness or after a late meal, so this page says nothing about either. A low night is not proof of illness, of overtraining or of any other cause.

If the number is not what you hoped, it may simply reflect normal night to night variation. The next section shows how much variation that can be.

How to follow your own trend

HRV changes through the night, and it is generally highest during the night and drops in the morning, so short readings are only comparable at the same time of day. The 2024 review notes that comparisons within one person are usually uncomplicated.⁶ The aim is to compare you with you.

1. Keep the method fixed. Use the same device and wear it the same way. In a validation study of wrist and ring devices, the average error in overnight HRV ranged from 5.96% for one device to 16.32% for another, and accuracy differed significantly between devices.⁷
2. Log a few things beside the number. Note alcohol, bedtime and how you feel each morning. This is a suggestion, so that you can look for a pattern later.
3. Change one habit at a time. If you change your bedtime and your drinking together, you cannot tell which did what.
4. Compare averages, not nights. Do not judge from one reading. A study by WHOOP authors found at least five of seven nights were needed to estimate a week of HRV variability reliably.¹⁶ Five of seven is 71% of the nights in a week.
5. Leave room for error. Nightly differences can be as large as the error of the device itself, as the example below shows.

The source for the five of seven nights is a commercial one: its authors are from WHOOP, a maker of wearables.¹⁶ It concerns how variable a week looks, not the level of HRV, and no source read gives a number of nights as a standard for setting a baseline. How best to set a baseline (fixed or rolling averages) still needs further study, according to a meta-analysis of HRV guided training, which also reported only small, non significant fitness gains for that approach compared with predefined training.¹⁷

A worked example (invented numbers)

Suppose seven nights read 54, 50, 56, 52, 49, 55 and 53 ms. The total is 369, so the average is 369 ÷ 7 = 52.7 ms. The gap between the highest and lowest nights is 56 − 49 = 7 ms, or 13.3% of the average. The lowest night, 49 ms, sits 7.0% below the average.

That is why a single night should not drive a decision, and why a pattern over a few weeks is more useful. For what a wrist device can and cannot do, see wrist versus chest strap heart rate. The British Heart Foundation describes wrist devices as light sensors measuring blood flow in the wrist, useful for spotting trends or changes, but unable to diagnose a heart problem.⁵

For how recovery scores combine sleep, resting heart rate and HRV, see the readiness score explained, the sleep score guide and sleep debt. Your pulse after exercise is a separate measure, covered in heart rate recovery, and your pulse at rest in the resting heart rate guide.

What HRV cannot tell you

• It cannot diagnose. The British Heart Foundation says wrist devices cannot diagnose a heart problem or give a full picture of heart health.⁵
• It has no agreed target. The sources read give no figure that is right for everyone, and a higher number is not automatically better.
• It does not prove a cause. The alcohol, exercise and stress findings are associations in groups of people, mostly from abstracts, and one night does not show which habit moved it.
• It does not travel between devices. Accuracy differed significantly between devices in the validation study, so compare readings from one device only.⁷

If a change in a number worries you, speak to your GP. The next section covers symptoms that matter far more than any HRV figure.

Frequently asked questions