What HRV and Sleep Scores Miss About Muscle Recovery
Recovery wearables made one number famous. Heart rate variability tells you whether your nervous system is rested. Sleep stages tell you whether your night was useful. Readiness scores tell you whether to go hard or back off.
None of them can tell you which muscle is still under-firing.
That gap matters most where recovery is the whole job. A physical therapist watching a patient come back from an ACL repair, a rotator cuff, or a chronic hamstring strain is not trying to score the patient's night. They are trying to answer one question: is this muscle, this side, this movement, ready for the next dose of load?
That question has an objective answer. It just happens to live in a signal that wrist wearables cannot read.
The signal recovery wearables cannot see
Surface EMG measures the electrical activity a muscle produces when it contracts. The amplitude scales with the size of the contraction. The timing tells you when the muscle is firing relative to the rest of the chain. The asymmetry between left and right tells you whether the recovering side is pulling its weight, or whether something else is compensating.
PPG sensors on a wrist cannot do any of this. Neither can a chest strap, a ring, or a strap on the bicep. They are reading blood-flow proxies for heart rate. Useful for cardio load and sleep. Mute on muscle work.
This is not a knock on the recovery category. It is a category line. WHOOP, Oura, and Garmin are general-purpose health trackers. They will not tell a clinician whether a patient's left vastus medialis is firing in time with the right one on day 30 of a quad rehab. That is not what they were built to do.
What recovery looks like when you can actually see the muscle
Three things become measurable that were not before.
1. Activation asymmetry. Two muscles of the same name on opposite sides should produce comparable activation during a comparable contraction. After unilateral injury, they usually do not. The size of that gap, and how it closes session over session, is one of the cleanest indicators of true neuromuscular recovery. It is also one of the easiest things to miss without direct measurement, because patients compensate. They will load the strong side without realizing it.
2. Onset timing. In coordinated movement, muscles fire in a sequence. Glute-then-hamstring, scapular stabilizer-then-deltoid, transverse abdominis-then-prime mover. After injury, the sequence frequently disorganizes — the prime mover fires first, the stabilizer comes late, or not at all. EMG shows that re-ordering directly. Once you can see it, you can train it back.
3. Fatigue pattern. Fresh muscle and tired muscle produce different EMG signatures. As a muscle fatigues, the median frequency of the signal shifts and amplitude changes. A patient three sets into a rehab session who is now firing inefficiently is not the same patient who started the session strong. That is useful information for dosing the next set, and for deciding when to stop.
None of these are diagnostic claims about the patient's underlying condition. They are objective measurements of muscle behavior in the moment — the kind of feedback that has historically required a research lab or a clinic-grade sEMG system, and the kind that fits awkwardly into a 30-minute clinic visit.
Why this matters for the practitioner before the patient
The cleanest use of a new measurement tool is on yourself first. Run a session, watch the data, find out what the signal looks like when a muscle is fresh, what it looks like when it is taxed, and what it looks like when it is recovering between sets.
For a PT, that personal baseline does a few things. It teaches the tool fast — much faster than reading the manual. It surfaces the questions a patient will eventually ask. And it produces a perspective the patient is going to value: a clinician who can describe what their own quad looks like coming back from a session is a clinician who can read what the patient's quad is doing in the room.
This is also why the Inara purchase is single-unit, single-user. The product is built for one person at a time, not for a clinic-wide dashboard. A practitioner who wants to use it with patients ends up either acquiring more sensors as the workflow scales, or — more commonly — having patients buy their own. Both are downstream effects of personal use, not features built into the product.
Where this fits in a recovery workflow
A few concrete patterns we have seen used:
- Asymmetry tracking session-over-session on the same lift or exercise — bridge variants for hamstrings, step-downs for quads, scapular work for shoulder. The gap closes, plateaus, or widens. The clinician dose-adjusts accordingly.
- Onset-sequence biofeedback during a movement re-education set. The patient gets a live visual of which muscle is firing and when. Verbal cues land harder when the patient can see what "fire the glute first" actually looks like in their own data.
- Fatigue-based set termination. Instead of ending a set at a fixed rep count, end it when the targeted muscle drops below a threshold of clean activation. Stops the compensating muscle from carrying the work the recovering one was supposed to do.
None of these replace clinical judgment. They make the judgment auditable.
The honest limits
Surface EMG is not magic. The signal is influenced by electrode placement, skin condition, subcutaneous tissue, and motion artifact. Comparing absolute amplitudes between sessions is harder than comparing within-session ratios. Asymmetry trends are more reliable than single-session absolute reads. Anyone selling EMG as a turnkey replacement for clinical assessment is overselling it.
What it is, reliably: a fast, objective, in-the-room measurement of which muscle is doing the work, and how that picture changes as the patient recovers. That is something HRV and sleep scores cannot give you, no matter how good the wrist sensor gets.
If you want to see what this looks like on yourself first
The Inara sensor is built for exactly this — single-user, on-body EMG for training and recovery sessions, with an iOS app that visualizes activation in real time. It is a coaching and feedback tool, not a diagnostic device. Most of our PT-side users start by running it on themselves for a few weeks before they ever use it with a patient.
- Read the physical therapy use cases — claims and workflows are written for US-licensed PTs.
- See how the sensor measures activation for the underlying technical picture.
The CTA is the same on every page: buy one sensor, use it on yourself, see what the signal actually looks like before you decide whether it has a place in your clinic workflow.
Inara is a wearable surface EMG biofeedback sensor for training and coaching feedback. It is not a diagnostic device and does not treat, cure, or prevent any condition. US physical therapists using Inara should apply it within their licensed scope of practice.