Sometime around 2018, the conversation about wearables in mental health shifted from "is this possible?" to "how do we deal with all this data?" A modern Apple Watch generates thousands of data points per day. An Oura ring measures heart rate variability every five minutes through the night. Fitbit silently tracks sleep stages, step counts, and resting heart rate every twenty-four hours.
For schizophrenia, the question is whether any of this matters. The honest answer is: parts of it do, parts of it don't, and the field is still figuring out which is which.
Sleep duration tracking, gross activity tracking, and very early relapse detection (in research settings) all show real signal. Direct prediction of psychosis or hallucinations from a wristband alone is still well outside what consumer devices can do today.
What wearables are good at
Sleep duration and timing
Of everything a wearable measures, sleep is the most clinically useful for schizophrenia. Sleep disruption is one of the strongest early warning signs of relapse — a person who suddenly drops from seven hours a night to four, or whose bedtime drifts steadily later over two weeks, is showing a pattern that a clinician would want to know about. Wrist-worn devices estimate sleep duration with reasonable accuracy compared to gold-standard polysomnography in healthy adults; the picture is messier in people with severe mental illness, but the broad signal still holds. See sleep hygiene for why this matters so much.
Activity and routine
Step counts and activity windows are objective measures of daily routine. A long, gradual reduction in daytime activity often correlates with worsening negative symptoms or depression. Conversely, sudden bursts of overnight activity can flag escalating agitation. Neither pattern is diagnostic, but both can prompt a useful conversation.
Heart rate at rest
Resting heart rate is a crude but real indicator of physical and mental load. Persistent elevation can reflect anxiety, dehydration, fever, alcohol use, or — relevant to people on antipsychotics — clozapine-induced tachycardia. Clozapine side effects include a baseline heart rate roughly 15–25 beats above pre-treatment, and a wearable trend line is one of the simplest ways to spot a sudden rise that warrants checking with a prescriber.
What wearables are not good at
Detecting hallucinations or delusions
Despite some optimistic press, no consumer wearable currently detects positive psychotic symptoms directly. Research groups are exploring patterns in voice, typing, and movement that may correlate with episodes, but the signal is not strong enough or specific enough for product-grade use, and there are major false-positive concerns.
Stress in any sophisticated way
Most "stress scores" on consumer devices are derived from heart rate variability (HRV). HRV does decrease with sympathetic activation, but it is also affected by sleep, alcohol, caffeine, exercise, illness, and many medications. Treating a single low-HRV day as evidence of psychological stress will mislead you most of the time.
Diagnosing relapse
The closest current research comes is from work by John Torous and colleagues at Beth Israel Deaconess (their 2020 paper in npj Digital Medicine is a good summary), which combined passive sensing with active patient-reported data to predict relapse with modest but real accuracy. None of this is yet packaged into a consumer product that says "you're relapsing." It is research, not a feature.
The harder questions
Adherence and engagement
Wearables only work if they are worn. A lot of the published research has high dropout rates — people stop charging the device, lose it, or simply find the constant tracking unpleasant. For some patients, particularly those with paranoid features, the sensation of being monitored by a device on their wrist can become part of a delusion. That is a real risk to weigh.
Data overload
Generating data is the easy part; using it well is the hard part. A trend graph that no one looks at is just a graph. The most successful uses pair the data with a person who can interpret it — a care coordinator, a family member, a clinician who reviews summaries between visits.
Equity
A $400 watch is a barrier. Schizophrenia disproportionately affects people who cannot easily afford the latest hardware, and the cheaper devices (basic Fitbits, older smartwatches) often do most of the same job for under $100. The best digital health tools are the ones that work on the device the user already has.
What we recommend if you're starting
- Pick the cheapest device that tracks sleep and steps. A basic Fitbit or an older Apple Watch is plenty.
- Look at one or two metrics, not twenty. Sleep duration and average daily steps tell you a lot. Stress scores and "readiness" rings tell you very little.
- Track for at least four weeks before drawing conclusions. A single weird day is noise; a two-week downward trend is signal.
- Share the trends with one person. A clinician, a parent, a partner. Data without a witness is just storage.
- Stop wearing it if it bothers you. No feature is worth daily distress.
Where the field is going
The serious research direction is "digital phenotyping" — combining passive sensor data (motion, sleep, screen time) with brief active check-ins to build a personalised model of stability for each individual. The most promising work uses smartphones, not wearables, because everyone already has a phone. Frida's own approach falls broadly into this category; we describe how in how Frida tracks cognitive stability.
Bottom line: wearables are a useful adjunct, not a treatment. They will not replace medication, therapy, or honest conversations with a clinician. But for people who want to see their own patterns and share them with a trusted person, they can quietly add real value.
This article is for educational purposes only and is not medical advice, diagnosis, or treatment. Always consult a qualified mental health professional. If you or someone you know is in crisis, call or text 988 in the US, or your local emergency number.