Why We Don't Show You a 5-Stage Sleep Hypnogram (And What We Show You Instead)
If you've used other sleep trackers, you've probably seen a colorful chart breaking your night into Wake, Light, Deep, and REM stages. It looks impressive. But here's the thing: most of those charts are guesswork — and even when they're accurate, they're not actually that useful.
At NextSense, we took a different approach. One informed by 18 months of work with the Digital Medicine Society (DiMe), the world's leading authority on digital health measurement standards.
What DiMe Taught Us (And What We Helped Teach DiMe)
In 2023, NextSense joined DiMe's Core Digital Measures of Sleep project as a founding contributor. Alongside researchers from Bayer, Lilly, Takeda, Jazz Pharmaceuticals, the VA, the NIH, and the FDA, we spent over a year defining what actually matters when you measure sleep outside a lab.
The result was a peer-reviewed paper — "Digital Measures Development: Lessons Learned from an Expert Workshop Addressing Cross-Therapeutic Area Measures of Sleep" (Fromy, Kremliovsky, Mignot, Aloia, Berent, et al., Digital Biomarkers, 2024) — and a comprehensive framework that identified three core measures that matter most:
- Sleep Latency — How long it takes you to fall asleep
- Total Sleep Time — How much sleep you actually got
- Wake After Sleep Onset (WASO) — How much you woke up during the night
Notice what's not on that list? A five-stage sleep hypnogram.
Why Five Stages Can Be Misleading
Traditional polysomnography (PSG) in a sleep lab classifies your sleep into Wake, N1, N2, N3 (deep), and REM — five stages, scored by trained technicians reviewing 30-second epochs of multi-channel brain data. It's the gold standard, and it requires electrodes glued to your scalp, a technician watching you all night, and about $3,000.
Consumer wearables that claim to show you these five stages are typically using motion and heart rate to estimate them — not measure them. The difference matters. When you see "20% Deep Sleep" on your wrist, that number may have little correlation with what a sleep lab would find.
NextSense SmartBuds actually can classify sleep stages with clinical-grade accuracy (κ = 0.73, matching PSG performance). We have in-ear EEG — the same type of brain signal used in sleep labs. But we made a deliberate choice not to lead with a five-stage breakdown.
What We Show You Instead
Informed by the DiMe framework, we focus on the metrics that are:
- Clinically validated — backed by decades of sleep medicine research
- Actionable — you can actually do something about them
- Meaningful to you — they correlate with how you feel the next day
That means you see:
- How long it took you to fall asleep — so you know if your wind-down routine is working
- How much you actually slept — the number that matters most for health outcomes
- How fragmented your night was — because waking up five times for two minutes each is very different from sleeping straight through
- Your slow wave activity — this is where NextSense goes beyond the DiMe framework. We don't just read your deep sleep; we actively enhance it with closed-loop audio stimulation, and we show you the measurable boost
The Science Behind the Decision
The DiMe project brought together over 50 experts across 30+ organizations to answer a deceptively simple question: if you could only measure a few things about sleep, what should they be?
The consensus was clear. Sleep latency, total sleep time, and wake after sleep onset capture the dimensions of sleep that matter most for health, for clinical trials, and for the person lying in bed wondering why they're still tired.
Showing someone that they got "18% REM" doesn't help them sleep better tomorrow. Showing them that it took 45 minutes to fall asleep — and that their audio stimulation boosted deep sleep by 12% — does.
Clinical Grade, Consumer Simple
NextSense SmartBuds are a clinical-grade EEG disguised as premium music earbuds. We have the sensor capability to do everything a sleep lab does. But the DiMe work reinforced what we already believed: more data isn't always better data. The right data, presented clearly, is what changes behavior and improves outcomes.
That's why 75% of our active users see a measurable boost in slow-wave sleep activity. Not because we show them more charts — but because we give them the information that matters, paired with technology that actually does something about it.
NextSense CEO Jonathan Berent is a co-author of "Digital Measures Development: Lessons Learned from an Expert Workshop Addressing Cross-Therapeutic Area Measures of Sleep" (Digital Biomarkers, 2024), the foundational paper for DiMe's Core Digital Measures of Sleep framework. The full framework and resources are available at datacc.dimesociety.org.
