Wearable documentation review · Checked November 21, 2026
A wearable stress alert cannot settle an argument about how you feel. It can tell you that a device detected a physiological pattern. Turning that observation into “you were anxious at dinner” adds an explanation that the sensor did not establish.
This is a useful limit, not a reason to discard the feature. A graph can prompt a better question about your day. Trouble starts when its label becomes the answer. We compared current Oura, Garmin and Google Health documentation, then built a small observation sheet that keeps the device result separate from your own account.
Four features that should not share one column
| Feature | What the documentation describes | Boundary to preserve |
|---|---|---|
| Oura Daytime Stress | Personal physiological response using heart rate, HRV, movement and average body temperature | Awake, relatively inactive periods; it cannot identify the source of stress |
| Garmin Stress Level | HRV-based estimate on a 0 to 100 scale | High-activity periods can be unmeasurable; a daily summary differs from a current reading |
| Google Health Body responses | Sudden, rapid increases in autonomic activity on compatible devices | An event notification, not a report of overall perceived stress |
| Google Health Resilience | Daily capacity assessment drawing on responsiveness, exertion balance and sleep | Current help uses Optimal, Balanced and Low; this is separate from Body responses |
Oura explicitly distinguishes biometrics from emotions. Its four zones are Stressed, Engaged, Relaxed and Restored, relative to the wearer's baseline. The page describes updates every 15 minutes during eligible waking periods. Movement, nonwear and signal problems can leave gaps. An empty stretch therefore does not establish that the person felt calm.
Garmin separates resting values of 0 to 25 from low, medium and high stress bands above that. It also distinguishes the device's current stress value from the overall daily value in Garmin Connect. A watch and a phone can show different numbers because the views summarize different periods. That possibility should be checked before concluding that one sensor is faulty.
Google Health's current help says Body responses can follow situations including standing up, brief walks, eating or caffeine. An alert arrives roughly ten minutes later and can prompt a mood entry. Its Resilience feature answers a different, daily question. Older terminology should not be pasted into a current comparison without checking the app version and feature.
An observation, a feeling and a cause are three records
Imagine a fictional evening: a notification arrives at 7:10 p.m.; the person remembers feeling happy; dinner and a lively conversation occurred around 7 p.m. These observations can coexist. The notification does not invalidate the reported feeling, and the pleasant feeling does not prove the sensor malfunctioned.
There are several possible explanations, including a physiological response associated with something in that period, timing differences between detection and notification, or a measurement problem. The timeline alone does not identify which explanation is right. “Dinner caused stress” would be a causal claim, not a transcription of the screen.
Our worksheet gives those statements separate fields: device observation, displayed event time, notification time, reported feeling, context and alternative explanation. Its rows are blank. The fictional dinner above is a teaching example, not a Vora member's record or a test of any brand.
The quietest-looking day may be the least observed
Suppose a hypothetical graph has a long empty section during a workout. Coloring that gap green would invent a low-stress observation. Filling it red would invent a high-stress observation. Both moves conceal the same problem: the feature did not supply a usable value for that interval.
A comparison between days needs to retain coverage. Ask whether the device was worn, whether the relevant feature was available, and whether movement or another limitation prevented a reading. A total of observed “stressed minutes” without the amount of eligible observation time is easy to overinterpret. We are not claiming any reviewed app uses that flawed calculation; this is a check for people making their own summaries.
The same caution applies across brands. A Garmin daily number, an Oura zone timeline and a Google Body response count are different kinds of outputs. Converting them to a common percentage does not make them calibrated measures of the same thing.
Use the prompt without surrendering the interpretation
Start with a modest question: “What was happening around this reading?” Save the exact feature and time, then add what you remember. If an association seems to repeat, look for ordinary alternative explanations and missing intervals before changing a routine around it. An observational diary can help you formulate a question; it does not isolate a cause.
Do not let a reassuring graph dismiss persistent symptoms or a distressing experience. Equally, a high reading is not a diagnosis of anxiety or a verdict on a relationship, job or meal. Those conclusions require information that the graph does not contain.
Method and reusable record
This is a manually checked map of public product definitions, not a validation study or a ranking of accuracy. We did not compare devices against clinical measures, test their alerts or inspect private health records. Features and naming can change. The source pages define the reviewed scope; availability depends on the relevant device and service.
The CSV and JSON preserve those distinctions and include an unfilled observation template. The separate HRV field guide explains another reason that apparently similar inputs should not be merged casually.
Publisher disclosure: Vora combines proactive AI with wearable, food, workout and recovery context. Useful context should help someone ask a better question, not manufacture certainty about an emotion. This reference stands alone; our product features explain the app. Corrections: support@askvora.com.