You meditate every morning. Your Oura ring shows your HRV trending up. Meditation works, right?
Maybe. But probably not for the reason you think.
The largest meta-analysis ever done on meditation and HRV — 19 randomized controlled trials, 1,465 participants — found a non-significant effect.
Let that land for a moment. The wellness industry's most sacred cow, tested rigorously, failed to reach statistical significance.
The Meta-Analytic Verdict: g = 0.19 (That's Basically Nothing)
Brown et al. (2021) published in Psychosomatic Medicine the definitive analysis. They found an overall effect of Hedges' g = 0.38 — but that included one statistical outlier (Muthukrishnan 2016) that reported an implausible g = 3.22.
Remove that outlier and the effect drops to g = 0.19.
Still non-significant. The confidence interval crossed zero.
But here's where it gets really uncomfortable. When they looked at only low-bias studies — the ones with proper blinding, randomization, and active controls — the effect dropped to g = 0.044.
For context, that's roughly the effect of wearing a lucky charm.
A second meta-analysis (Bin Waleed et al. 2022, 21 studies, 538 meditators) confirmed the null finding across every standard HRV metric: SDNN, RMSSD, LF, HF — no significant differences between meditators and controls.
Better Studies → Smaller Effects
This is the most damning pattern in the data. Brown et al. found that risk of bias was a significant moderator (p = .041):
High-bias studies (poor methodology): g = 0.64 — looks impressive.
Low-bias studies (rigorous methodology): g = 0.044 — basically zero.
When meditation was compared to active controls (doing something else, like reading or walking): g = -0.03. Literally zero.
When compared to passive controls (doing nothing): g = 0.52. Seems like meditation works! But so would watching TV.
The "meditation improves HRV" narrative is largely an artifact of study quality.
Loving-Kindness Meditation Makes Your HRV Worse
Lumma, Kok, and Singer (2015) tracked about 80 participants through the ReSource Project, practicing three different meditation types daily for three months each.
Breathing meditation: lower heart rate, higher HF-HRV. Relaxing, as expected.
Loving-kindness meditation: higher heart rate, lower HF-HRV. Sympathetic activation. The opposite of what the wellness narrative predicts.
Observing-thoughts meditation: same sympathetic pattern as loving-kindness.
And over three months of practice, HF-HRV actually decreased further for loving-kindness. The more you practiced, the more aroused your nervous system became.
The study's title says it all: "Is Meditation Always Relaxing?" No. It is not.
The Beginner's Advantage (Experts Get Less)
This one contradicts every assumption about meditation mastery.
Bin Waleed et al. found that normalized HF was significantly higher DURING meditation in beginners but NOT in experienced meditators.
Chang, Chueh, and Lai (2020) confirmed it: beginners (n=18) showed continued HRV improvement over one month. Experienced practitioners (average 9 years, n=27) showed session-by-session improvements but zero additional gains over the same month.
Wei et al. (2025) added the explanation. Experienced meditators' HRV actually drops initially during meditation — a "rise-then-fall" pattern — while novices show monotonic increases. The expert pattern reflects "autonomic flexibility" but would look worse on your wearable during a session.
So if you've been meditating for years and your HRV hasn't budged, this is not a sign you're doing it wrong. You may have already captured whatever benefit exists.
Your Wearable Is Measuring the Wrong Metric
Natarajan (2023) analyzed PhysioNet data from seven cohorts and found something that should concern every wearable user.
RMSSD during Chi meditation: dropped from 36.4 to 35.5 ms. Effect size d = -0.1. Essentially zero.
Autonomic Balance Index during the same meditation: jumped from 0.19 to 0.72. Effect size d = 3.9. Massive.
Same meditation. Same person. Same session. One metric says nothing happened. The other says everything changed.
The problem: RMSSD preferentially weights higher-frequency vagal components and systematically misses the slow-breathing-induced parasympathetic shifts that meditation produces.
Every consumer wearable — Oura, Whoop, Garmin, Apple Watch — relies on RMSSD. The metric may be fundamentally wrong for evaluating meditation's autonomic effects.
84% of Studies Didn't Control for Breathing
This is the elephant in the meditation research room.
Only 3 of 19 RCTs in Brown et al.'s meta-analysis controlled for respiration rate.
Meditation naturally slows breathing. Slower breathing mechanically increases HRV independent of any vagal tone change. When you breathe at 6 breaths per minute, your HRV goes up — whether you're meditating, watching Netflix, or counting sheep.
Panjwani et al. (2019) demonstrated this directly. Heartfulness meditation actually reduced total HRV power. The apparent HF increase was an artifact of LF dropping more than HF. Their conclusion: the changes were "due solely to a breathing effect."
The rare exception: Nesvold et al. (2012) found that nondirective meditation (ACEM) increased HRV in both LF and HF bands without changing respiratory rate. This is one of the only studies showing a genuine meditation-specific — not breathing-driven — HRV effect. But the meditation type matters: nondirective means letting thoughts wander with minimal effort. The opposite of focused mindfulness.
Movement Meditation Crushes Seated Meditation (3-10x Effect)
If you want meditation to actually improve your HRV, stand up.
Zou et al. (2018) meta-analyzed 17 RCTs (~1,277 participants) of yoga and tai chi:
Yoga/tai chi: normalized HF Hedges' g = 0.37 (p < .001), LF/HF ratio g = -0.58 (p < .001). All significant.
Compare that to seated meditation's g = 0.044-0.19 (non-significant).
That's a 3-10x difference in effect size. And it actually reaches statistical significance.
About 60 minutes per week of yoga was the threshold for meaningful improvement. Brown et al. explicitly noted: "Physical activity-based interventions may offer a more plausible pathway to improved HRV than seated meditation practice."
No Dose-Response (And It Reverses at Follow-Up)
From Brown et al.: no dose-response relationship was found. Meditating more did NOT produce bigger effects. Longer interventions did NOT outperform shorter ones.
And at 23.5-week follow-up, the effect actually reversed to g = -0.18.
Whatever meditation does to HRV, it doesn't accumulate. And it doesn't stick.
The One Exception: Kirk and Axelsen (2020), in a 10-day Headspace RCT with 90 participants, found r = 0.47 between total practice minutes and RMSSD change. But one study versus a meta-analysis showing no dose-response is a weak argument.
The Harm Question: 8-33% Experience Adverse Effects
Farias et al. (2020) systematically reviewed 83 studies covering 6,703 meditation practitioners.
Overall adverse event rate: 8.3%. In observational studies (capturing real-world practice): 33.2%.
Most common: anxiety (33% of adverse events), depression (27%), cognitive anomalies (25%). These occurred even in people with no prior mental health history.
The population most attracted to meditation for HRV improvement — stressed, anxious, burnt-out — is the same population at highest risk.
What to Actually Do With This Information
This is not an argument against meditation. Meditation has genuine benefits for stress perception, emotional regulation, focus, and anxiety that don't require HRV improvement to be real.
But if you're meditating specifically because you want to see your resting HRV go up on your wearable, the evidence says:
1. Add deliberate slow breathing (5-6 breaths per minute). This has the strongest evidence of any autonomic intervention (Laborde 2022, 223 studies). It works whether or not you're meditating while doing it.
2. Consider yoga or tai chi instead of seated meditation. 3-10x larger effect sizes, and they actually reach statistical significance. 60 minutes per week is the threshold.
3. Don't judge meditation by your RMSSD score. The metric your wearable reports may be systematically blind to meditation's actual autonomic effects. Your meditation may be working in ways your Oura ring cannot detect.
4. If you're an experienced meditator seeing no HRV gains, that's normal. The beginner boost doesn't last, and the dose-response curve is flat. You're meditating for other reasons now. That's fine.
5. Try nondirective meditation if HRV specifically matters to you. It's the one type that showed genuine HRV improvement without the breathing confound. Let your mind wander. Don't focus. The irony: trying less produces more.
The uncomfortable truth for wearable enthusiasts: the thing you track most easily (HRV) may be the wrong lens for evaluating the thing you do most faithfully (meditation). The benefits are real. The measurement just can't see them.
Sources
Brown et al. (2021). Effects of Mindfulness and Meditation on HRV: A Meta-Analysis. Psychosomatic Medicine. N=1,465, 19 RCTs. PMID 33395216.
Bin Waleed et al. (2022). Meditation and Cardiac Autonomic Modulation. EP Europace. 21 studies, 538 meditators.
Lumma, Kok & Singer (2015). Is Meditation Always Relaxing? Int J Psychophysiology. ~80 participants. PMID 25937346.
Chang, Chueh & Lai (2020). Beginner vs Experienced Meditators. IJERPH. n=45. PMID 32210043.
Wei et al. (2025). EEG-ANS Dynamics in Experienced vs Novice Meditators. BMC Complementary Medicine. PMID 40611081.
Natarajan (2023). Autonomic Balance Index vs RMSSD. Frontiers in Physiology. 7 cohorts. PMID 36756034.
Panjwani et al. (2019). Heartfulness Meditation and ANS. Frontiers in Cardiovascular Medicine. PMC6527777.
Nesvold et al. (2012). Nondirective Meditation and Vagal Tone. European J Preventive Cardiology. PMID 21693507.
Zou et al. (2018). Yoga/Tai Chi and HRV. J Clinical Medicine. 17 RCTs, ~1,277 participants. PMID 30384420.
Kirk & Axelsen (2020). 10-Day Headspace RCT. n=90. PMID 33332403.
Farias et al. (2020). Adverse Effects of Meditation. Acta Psychiatrica Scandinavica. 83 studies, 6,703 participants. PMID 32820538.
Laborde et al. (2022). Slow-Paced Breathing and HRV. 223 studies. PMID 35623448.