The pitch for a meditation wearable is seductive. Sit still, breathe, and let a small device tell you when your brain or your heart lands in the right state. No guessing, no wondering whether you are doing it right. Just data and a nudge.
The reality is messier. Three devices dominate this category — the Muse EEG headband, the HeartMath emWave breathing trainer, and the Apollo Neuro vibrational wristband — and each one measures something different, rests on a different evidence base, and answers a different question. This piece walks through all three honestly, then names the thing none of the marketing pages will say out loud.
Three different questions
Before comparing devices, it helps to notice that they are not really competitors. They are answering different questions.
- Muse asks: is your brain in a calm attentional state right now? It uses electroencephalography (EEG) to sample brain activity in real time.
- HeartMath asks: are you breathing in a rhythm that maximises heart rate variability? It uses a pulse sensor to score how coherent your breath-linked heart rhythm is.
- Apollo Neuro asks: can gentle vibration on the skin shift you toward calm or focus? It does not measure anything about you. It emits patterned vibrations and hopes the nervous system responds.
Only the first two are true biofeedback devices — they read a signal from your body and give you back a representation of it in real time. Apollo Neuro is a stimulation device, not a feedback device. That distinction matters when you look at the evidence.
Muse — what an EEG headband actually sees
The Muse 2 is a thin plastic headband with five dry electrodes — two behind the ears and three across the forehead. Those electrodes pick up the small voltage fluctuations produced by cortical neurons firing in synchrony. The device streams the signal to your phone, where the Muse app classifies it into rough frequency bands: delta (deep sleep), theta (drowsy, meditative), alpha (relaxed wakefulness), and beta (alert, active thought).
During a guided meditation, the app plays weather sounds that mirror the classification. Windy weather means the algorithm thinks your mind is active. Calm weather means it thinks you have settled. The auditory feedback is subtle — it does not tell you what to do, it just changes as your state changes.
A few honest points about consumer EEG.
- It is real EEG, but low channel count. Clinical EEG uses 19 to 64 electrodes with conductive gel. Muse uses five dry electrodes. That is enough to detect broad shifts in frontal-lobe activity, not enough to localise a source or map fine-grained cognition.
- Motion and muscle artefact are the main enemies. Jaw tension, blinking, and small head movements all produce voltages that swamp cortical signals. The algorithm tries to filter these out, but the filtering is imperfect. If you fidget, your Muse session will look worse than it is.
- Alpha and theta are correlated with meditative states, not identical to them. An experienced meditator can be deeply focused with a beta-heavy signal. A beginner can produce a lot of alpha simply by closing their eyes.
What the research actually shows
The evidence base for Muse is modest but real. A frequently-cited feasibility study by Bhayee and colleagues (2016) trained novice users with Muse over six weeks and measured improvements on standard cognitive attention tasks and self-reported mood. Later randomised work has tested Muse-supported meditation against unguided practice for anxiety, sleep, and workplace stress. Effect sizes are small to moderate, sample sizes are small, and — critically — the comparison is usually Muse-plus-meditation versus no meditation, not Muse-guided meditation versus unguided meditation.
Translation: the evidence tells us that people who use Muse tend to meditate more consistently, and people who meditate consistently tend to feel better. Whether the neurofeedback itself adds value on top of doing the practice is less clear.
Muse earns a 70 Standard in our system. The Standard tier means we can verify the device does what it claims to do at a hardware level — the EEG is real, the sensor manufacturer is transparent, the app is well-built — but the outcome evidence is not strong enough to push it into the Verified band alongside FDA-cleared cardiac wearables. That is a fair placement.
HeartMath emWave — the case for coherent breathing
The HeartMath emWave devices are not in our catalog, but the underlying practice is worth understanding because it is closer to universal than most.
Every mammal shows an effect called respiratory sinus arrhythmia (RSA): heart rate rises slightly on inhalation and falls slightly on exhalation. In a healthy young adult, the swing can be substantial. In an ageing or stressed nervous system, it flattens. RSA is one of the components clinicians pool into overall heart rate variability, or HRV.
Here is the interesting part. When people breathe slowly and evenly — around six breaths per minute, or 0.1 Hz — the amplitude of RSA increases dramatically. The heart rhythm and breath rhythm lock into a resonance pattern. That resonance is what HeartMath calls "coherence." It is a real physiological state. Under six-breath-per-minute breathing, the parasympathetic nervous system dominates, blood pressure gently oscillates in phase with breath, and subjective calm typically follows within a couple of minutes.
The emWave devices are pulse sensors — earlobe clip or fingertip — that display a real-time coherence score. When your breath and heart lock in, the score climbs. That is genuine biofeedback for a real physiological target.
The published work here is more mixed than HeartMath marketing implies. Independent reviews of resonant-frequency breathing (not tied to a specific device) have found meaningful reductions in blood pressure and anxiety scores in some populations. Reviews of the HeartMath training programme itself, especially in workplace and clinical settings, show smaller and more variable effects. Some of the strongest positive studies are co-authored by the HeartMath Institute, which is a conflict of interest worth noting.
The honest interpretation is this: slow paced breathing at roughly six breaths per minute works. You do not need a $200 device to do it. A metronome app or a simple visual pacer — free — will get you the same physiological outcome. The device is useful mainly if watching a coherence score keeps you at it longer.
Apollo Neuro — the weakest evidence base of the three
The Apollo Neuro is a wristband or ankle band that vibrates in patterned rhythms. Different vibration patterns are marketed for different goals — energy, focus, calm, sleep. The company argues that low-frequency vibration on the skin is picked up by mechanoreceptors and modulates the autonomic nervous system, shifting HRV and self-reported state.
Here is the problem. Apollo does not measure anything about you. There is no sensor reading your HRV, breath, or brain activity. It vibrates on a schedule. That is not biofeedback; that is timed haptic stimulation, and the mechanism by which it would reliably shift autonomic state is not well established outside of the company's own materials.
The published evidence is thin and heavily company-linked. A small randomised trial funded by Apollo (Siepmann et al., 2022, and adjacent work from the University of Pittsburgh group with Apollo funding disclosures) reported HRV improvements during device wear. Independent replication is limited. There is no FDA clearance for any therapeutic claim; the device is sold as a general wellness product.
None of this means Apollo Neuro does nothing. Gentle rhythmic stimulation may help some users relax the same way a weighted blanket does, and the placebo response to any deliberately calming ritual is real. It does mean that in the current published record, Apollo has substantially weaker evidence than either Muse or paced-breathing biofeedback, and its price point (typically $349 and up) is hard to justify on the data alone.
The uncomfortable truth about all three
The largest, most replicated, and most clinically meaningful finding across the entire meditation literature is not about any device. It is about consistency.
Studies of mindfulness-based stress reduction (MBSR), transcendental meditation, and open-monitoring practice consistently find that outcomes correlate with total hours of practice and with regularity — daily beats weekly beats sporadic — far more than with any specific technique or aid. A person who meditates ten minutes a day, unguided, for a year will likely see more benefit than a person who buys a $500 device and uses it three times.
This is why we score meditation wearables the way we do. Muse is a well-built EEG platform that helps some people build a daily habit. HeartMath and its imitators point at a real physiological effect (paced breathing) that you can access for free. Apollo Neuro asks you to trust a mechanism the published record has not clearly demonstrated. None of them will make you a meditator. Only sitting down every day will do that.
How to actually use one
If you decide the feedback is a useful crutch, here is how to get real value from it.
- Commit to a fixed sit time before you buy the device. Ten minutes a day, six days a week, for four weeks. Put it in the calendar. If you cannot do that unassisted, the device will not save you. If you can, it may sharpen what you are already doing.
- Ignore the score during the sit. Watching the meter in real time turns meditation into a video game and pulls you out of the practice. Sit with eyes closed. Check the summary afterwards.
- Watch the four-week trend, not the daily number. One session's score reflects noise, artefact, and how tired you were more than skill. A rising four-week average of calm-state time or coherence percentage reflects something real.
- Correlate with your day, not with a meditation score. The question that matters is whether the practice reduces your reactivity, improves your sleep, or increases your working memory. Track those. The device metric is a proxy.
Who each device is actually for
- Muse 2 — Best for people who want a concrete signal that meditation is producing a state change, and who like the feedback loop of a passive skill-building game. Weakest for people who already have a stable daily practice.
- HeartMath emWave — Best for people whose specific goal is autonomic regulation (blood pressure, anxiety, panic) and who want to see a live coherence readout. Free alternatives exist. The device pays for itself only if it makes you actually practice.
- Apollo Neuro — At current evidence, we would not recommend it as a first purchase. The mechanism is speculative, most supportive studies are company-funded, and the price is high. Try a weighted blanket or a paced breathing app first.
The bottom line
A meditation wearable is a training aid, not a shortcut. The Muse 2 is the only device in this category that clears our Standard bar for hardware honesty and modest published evidence, and it earns its 70 Standard score by helping some users build a habit that would otherwise dissolve. HeartMath points at a real physiological lever — paced breathing at roughly six breaths per minute — that you can pull without any device at all. Apollo Neuro is currently sold on a mechanism the independent literature has not yet demonstrated.
If you already meditate and you want a tool to sharpen your practice, Muse is defensible. If you have never sat still for ten minutes, buy a timer, close your eyes, and prove to yourself you will show up. Any of these devices will only ever amplify a habit you already have — and the biggest predictor of benefit, across the entire published literature, is the one that costs nothing.