You wear a smartwatch to count steps and track sleep. But what if that data could also warn you about your mental health – before you even notice a problem? A study published in Scientific Reports (Nature Portfolio) has done exactly that. Researchers in Japan equipped 81 workers with Fitbit Sense 2 devices and single-channel sleep EEG monitors for five days. They then correlated objective sleep and activity data with validated mental health questionnaires measuring distress, anxiety, depression, and personality traits.
The findings are clear, objective, and actionable: Your sleep patterns and daily activity levels are directly linked to your mental health – and wearable devices can detect these signals.
The Headline: Sleep Timing and Deep Sleep Matter Most
The study found that specific, measurable sleep parameters were strongly associated with mental health outcomes:
- Later bedtimes correlated with higher state anxiety (ρ = 0.30, p = 0.024). This relationship was independent of total sleep time – meaning even if you get enough hours, going to bed late is linked to worse anxiety.
- Longer sleep onset latency (time to fall asleep) was associated with higher distress severity (ρ = 0.28, p = 0.034) and greater interference of distress with daily life (ρ = 0.33, p = 0.010).
- Lower sleep efficiency (more time awake in bed) correlated with higher distress interference (ρ = -0.31, p = 0.017).
- Decreased N2 sleep (light non-REM sleep) was linked to higher anxiety (ρ = -0.29, p = 0.028) and depression scores (ρ = -0.26, p = 0.049).
- Increased N3 deep sleep was associated with lower harm avoidance – a personality trait linked to chronic worrying and risk aversion (ρ = -0.39, p = 0.002). More deep sleep means less chronic anxiety.
Key Sleep Parameters Linked to Mental Health
| Sleep Metric | Mental Health Association | Strength (ρ) | Significance |
|---|---|---|---|
| Later bedtime | Higher state anxiety | 0.30 | p = 0.024 |
| Longer sleep onset latency | Higher distress severity | 0.28 | p = 0.034 |
| Longer sleep onset latency | Higher distress impact | 0.33 | p = 0.010 |
| Lower sleep efficiency | Higher distress impact | -0.31 | p = 0.017 |
| Decreased N2 sleep | Higher anxiety & depression | -0.29, -0.26 | p < 0.05 |
| Increased N3 deep sleep | Lower harm avoidance | -0.39 | p = 0.002 |
| Longer total sleep time | Lower insomnia severity | -0.26 | p = 0.043 |
ρ = Spearman’s rank correlation coefficient. Negative ρ for protective factors (e.g., more deep sleep = less harm avoidance).
The Activity Finding: Vigorous Exercise Lowers Harm Avoidance
Beyond sleep, the study found that vigorous-intensity daily activity (high-intensity exercise that raises heart rate significantly) was associated with lower harm avoidance scores (ρ = -0.35, p = 0.002).
Harm avoidance is a personality trait characterized by:
- Anticipatory worry
- Fear of uncertainty
- Shyness toward strangers
- Fatigability
People with high harm avoidance are more likely to be anxious, risk-averse, and prone to chronic worry. The study suggests that regular vigorous activity may reduce this trait, potentially protecting against anxiety disorders.
Practical takeaway: High-intensity exercise – not just walking – appears to have unique mental health benefits. This aligns with the WHO recommendation of 75 minutes of vigorous activity per week.
The Subjective Factor: Restorativeness Matters More Than You Think
One of the most striking findings involved sleep restorativeness – how refreshed you feel after sleeping. Participants rated their sleep quality on a 1-10 scale each morning.
Restorativeness showed no significant correlation with any objective sleep EEG metric. That is, how refreshed you feel does not necessarily match what your brain waves are doing.
But here is the critical part: Restorativeness was strongly correlated with every mental health scale:
- Depression (PHQ-9): ρ = -0.62 (p < 0.0001)
- State anxiety (STAI-S): ρ = -0.49 (p < 0.0001)
- Distress severity: ρ = -0.42 (p < 0.0001)
- Distress impact: ρ = -0.48 (p < 0.0001)
- Insomnia severity: ρ = -0.45 (p < 0.0001)
- Poor sleep quality (PSQI): ρ = -0.45 (p < 0.0001)
- Hyperarousal: ρ = -0.36 (p = 0.001)
Sleep Restorativeness – Strongest Mental Health Correlations
| Mental Health Measure | Correlation with Restorativeness (ρ) | Interpretation |
|---|---|---|
| Depression (PHQ-9) | -0.62 | Lower restorativeness = higher depression |
| State Anxiety (STAI-S) | -0.49 | Lower restorativeness = higher anxiety |
| Distress Impact | -0.48 | Lower restorativeness = more life interference |
| Insomnia Severity (ISI) | -0.45 | Lower restorativeness = worse insomnia |
| Poor Sleep Quality (PSQI) | -0.45 | Lower restorativeness = poorer sleep |
| Hyperarousal (HAS) | -0.36 | Lower restorativeness = more hyperarousal |
*All p < 0.001 except hyperarousal (p = 0.001).*
What this means: You can have objectively “normal” sleep (normal duration, normal stages) but still feel unrefreshed. And that subjective feeling of unrefreshing sleep is a powerful signal of depression, anxiety, and distress. If you wake up tired every day, pay attention – your mental health may need support.
Why This Matters for the Workplace
The study was conducted on Japanese workers during the COVID-19 pandemic (2020-2024), a period of significant stress and routine disruption. However, the findings have universal implications:
- Wearables can screen for mental health risk. Instead of relying on self-reported sleep diaries (which are often inaccurate), companies and healthcare providers could use anonymized wearable data to identify employees at risk for depression, anxiety, or insomnia.
- Late bedtimes are a red flag. Even if total sleep time is adequate, going to bed late (shifted circadian rhythm) is linked to higher anxiety. Workplace policies that respect natural sleep rhythms (e.g., flexible start times) could reduce mental health burden.
- Deep sleep is protective. Deep N3 sleep was strongly associated with lower harm avoidance (chronic worrying). Interventions that improve deep sleep – such as exercise, stress reduction, and consistent bedtimes – may also reduce anxiety.
- Subjective restorativeness cannot be ignored. Asking employees a single question (“How refreshed do you feel?”) may be as informative as complex sleep tracking. Low restorativeness scores should trigger mental health screening.
The Science: Why Sleep Stages Affect Mental Health
The study found that N2 sleep (light non-REM sleep) was reduced in people with higher anxiety and depression, while N3 deep sleep was increased in people with lower harm avoidance.
This aligns with a large body of research:
- N2 sleep is involved in memory consolidation and emotional regulation. Disrupted N2 is common in mood disorders.
- N3 deep sleep is the most restorative stage. It lowers cortisol, reduces sympathetic nervous system activity, and promotes emotional stability.
Practical takeaway: Improving sleep quality is not just about getting 7-8 hours. It is about protecting deep sleep. Avoiding alcohol before bed, reducing evening light exposure, and maintaining a cool bedroom temperature all promote N3 sleep.
The Limitations (Honest and Transparent)
The study has several limitations:
- Cross-sectional design: Cannot prove causation. It is possible that poor mental health causes poor sleep, rather than the reverse (likely bidirectional).
- Small sample size: Only 81 participants, with usable EEG data from 56. Larger studies are needed.
- Single-night EEG: Sleep patterns vary night to night. A single recording may not represent habitual sleep.
- Specific population: Japanese workers during COVID-19. Results may differ in other cultures or time periods.
- No pre-pandemic baseline: Cannot attribute findings specifically to the pandemic.
However, the strong, consistent associations across multiple sleep metrics and mental health scales suggest real, meaningful relationships.
Practical Recommendations for You
Based on this study, here is your evidence-based action plan:
- Pay attention to bedtime, not just sleep duration. Going to bed after midnight? Even if you sleep 8 hours, your anxiety risk may be higher. Aim for consistent, earlier bedtimes.
- If you take more than 30 minutes to fall asleep regularly, talk to a doctor. Long sleep onset latency is a strong signal of distress.
- Prioritize deep sleep. Reduce alcohol, caffeine, and screen time before bed. Exercise during the day. Keep your bedroom cool (65-68°F / 18-20°C).
- Do vigorous exercise at least twice a week. High-intensity activity (running, swimming, HIIT) was uniquely associated with lower harm avoidance. Walking is good, but intensity matters.
- Track your restorativeness. Each morning, rate how refreshed you feel (1-10). If your score is consistently low (below 5), seek mental health screening even if you think you slept “enough.” Your subjective experience is valid and informative.
The Bottom Line
This study provides objective, wearable-based evidence that sleep and activity patterns are directly linked to mental health. Longer sleep onset latency, later bedtimes, reduced deep sleep, and low vigorous activity are all associated with worse outcomes.
But here is the hopeful part: These are modifiable factors. You can change your bedtime. You can add vigorous exercise. You can improve sleep hygiene to increase deep sleep. And as you do, your mental health may improve alongside.
Your wearable is not just counting steps. It may be quietly monitoring your mental health. Listen to what it is telling you.