Imagine your wristwatch tapping you on the shoulder and saying, "Hey, tomorrow might be rough. Let's get you some help." That's not science fiction. A new study suggests we might be closer to that reality than most people think - especially for those who need it most.

Suicide risk doesn't follow a neat schedule. It spikes, dips, and shifts sometimes within hours. For military service members and veterans, who face disproportionately high rates of suicidal ideation and attempts, those fluctuations can be life-or-death. The challenge has always been catching the storm before it hits.

A team of researchers recently asked a simple but powerful question: can your sleep last night predict how you'll feel tomorrow? And can a Fitbit help answer that?

The Setup: Phones, Fitbits, and 28 Days of Data

The study enrolled 86 military service members and veterans, all of whom were experiencing current suicidal ideation or had attempted suicide within the past month. This is not a mild-risk sample. These are people in acute distress.

For 28 consecutive days, participants did two things. First, they answered seven brief surveys per day through their phones - a method called ecological momentary assessment, or EMA. Think of it as micro-journaling. Quick check-ins about sleep quality, nightmares, mood, and suicidal urges, collected in the flow of daily life.

Second, they wore a Fitbit. The wearable passively tracked sleep duration, sleep regularity, and other metrics without requiring any effort from the wearer. No buttons to push. No questions to answer. Just wear the thing and go about your day.

The researchers then pitted three approaches against each other: EMA-only, wearable-only, and a combined model using both. The goal was to see which best predicted next-day suicidal urges, depression symptoms, and PTSD symptoms.

What They Found: Nightmares Are a Red Flag

The results were striking, if perhaps unsurprising to anyone who's ever had a truly terrible night of sleep. More severe nightmares and poorer self-reported sleep quality - both captured through EMA - predicted next-day suicidal urges, suicidal beliefs, depression, and PTSD symptoms. Across the board.

Nightmares, it turns out, are not just unpleasant background noise. They're signal flares.

The Fitbit data told a different but complementary story. Deviations in sleep duration - sleeping significantly more or less than your personal average - predicted next-day maximum suicidal urges. Meanwhile, the wearable's sleep regularity index (basically how consistent your sleep schedule is) predicted next-day depression.

Here's where it gets interesting from a practical standpoint. When the researchers compared models head to head, EMA-only consistently outperformed wearable-only in predicting next-day outcomes. Your own words about your sleep were more informative than what the device measured on your wrist. But the best performance came from combining both approaches. Self-report plus passive sensing beat either alone.

Why This Matters: The Case for "Just-in-Time" Interventions

The mental health field has been chasing a concept called "just-in-time adaptive interventions," or JITAIs. The idea is beautifully simple. Instead of waiting for a weekly therapy appointment to address a crisis that happened on Tuesday, you deliver the right support at the right moment. Maybe it's a text. Maybe it's a coping exercise pushed to your phone. Maybe it's a prompt to call your therapist or a crisis line.

But JITAIs need fuel. They need real-time data that actually predicts when someone is heading toward danger. This study provides evidence that sleep data - both self-reported and passively sensed - could be that fuel.

Consider the implications. A veteran goes to bed. Their Fitbit registers a wildly irregular sleep pattern compared to their baseline. Their morning EMA check-in flags severe nightmares and poor sleep quality. A system could flag this combination and deploy an intervention before the worst of the next day unfolds.

We're not talking about replacing therapists with algorithms. We're talking about filling the gaps between appointments with something smarter than silence.

The Caveats (Because There Are Always Caveats)

A few things worth noting. The study had 86 participants. That's a meaningful sample for this type of intensive longitudinal design, but it's not thousands. Replication in larger, more diverse populations will be necessary.

EMA adherence was 48.9%. That means participants answered roughly half of the surveys they were prompted to complete. That's actually decent for seven-surveys-a-day-for-a-month, especially in a sample experiencing significant mental health distress. But it does raise the question: in a real-world deployment, would people keep responding? Fitbit adherence was better at 79.4%, which speaks to the appeal of passive sensing. It demands nothing from you.

There's also the self-report paradox. EMA outperformed the wearable, but EMA requires effort. The very people most at risk might be the least likely to answer their phone at 2 PM on a bad day. Wearables, while less predictive on their own, capture data even when someone can't or won't engage. That low-burden characteristic has real value, particularly for populations already burdened by their conditions.

Looking Ahead: Sleep as a Window Into Tomorrow

This study adds to a growing body of work suggesting that sleep is not just a consequence of poor mental health - it's a leading indicator. What happens at night shapes what happens the next day. For veterans and service members navigating PTSD, depression, and suicidal ideation, the relationship between sleep and next-day risk appears robust enough to build upon.

The combination of asking people how they slept and measuring how they slept creates a richer picture than either approach alone. That finding has direct engineering implications for anyone building digital mental health tools.

We're still in the early innings of this work. But the trajectory is promising. Your sleep data, both the kind you report and the kind your wrist quietly collects, might one day serve as an early warning system. Not perfect. Not a replacement for human care. But a bridge across the dangerous gaps in between.

And honestly, if your Fitbit can already judge you for not hitting 10,000 steps, it might as well do something actually useful.

This blog post discusses research findings and should not be taken as medical advice. If you or someone you know is experiencing suicidal thoughts, please contact the 988 Suicide and Crisis Lifeline by calling or texting 988. Research discussed here represents ongoing scientific investigation and clinical validation is still in progress.

All images used in this post are decorative illustrations only and do not represent or reflect the accuracy, reality, or correctness of the referenced research.

Primary Source: Predicting next-day suicidal urges, depression, and PTSD symptoms using ecological momentary assessments and passive sensing of sleep among military service members and veterans. PubMed. 2026. PMID: 41895610