Your nights arrive by themselves. YUTHI calculates how regular your sleep is — and after a short night, searches your entries for what's behind it.
The app calculates three things from your nights and your day — the sources are listed at the bottom of the page.
In a study of more than 60 000 people, regularity predicted mortality risk better than sleep duration — a connection, not a cause.5 The app calculates it for you as a percentage.
Each dot is a time you fell asleep. The app combines falling asleep and getting up: 100% would mean the same time every night; each hour of average deviation costs about 33 points.
If a night was shorter than you need, the app first searches the hours before for the cause itself. Only if it finds nothing does it ask you.
What the detective checks
Little daylight only counts if you usually log sun. Plus alcohol in the 8 hours before sleep and going to bed at least an hour later than usual. Then it compares: “On evenings with afternoon caffeine you sleep … minutes less on average.” If it finds nothing, it asks about your phone in bed, a room that's too warm or noise.
Cortisol rises clearly in the first 30 to 45 minutes after waking and then falls over the day.8 The app aligns the curve with the time you get up.
The curve is a schema, not a measurement. The points don't change it — they only show what happened in your day. If you have a cortisol blood value, it appears next to it.
From your bedtime, the app calculates when a warm shower fits and when your last coffee should be. In an analysis of 17 studies, a warm shower or bath 1 to 2 hours before sleep shortened the time to fall asleep.6
With a watch and Apple Health or with Oura, every night lands in the app in the morning, sleep stages included. YUTHI never overwrites what you log by hand.
Wrist devices detect sleep well, but individual stages only roughly — look at the trend, not a single night.9
An expert panel recommends 7 to 9 hours for adults and 7 to 8 for older adults.4 The app compares your average with your need and tells you how many minutes earlier you could go to bed.
One bad night is normal. A series of them shows a pattern — that's why the average over weeks counts.
If you work shifts, the app tells you honestly: the US sleep medicine guideline only gives conditional recommendations here, and the evidence is often thin.
A clockwise shift order is better than the reverse. Bright light during the night shift and caffeine before or during the shift can help against tiredness — but not shortly before sleep.10
With a watch, you see when you were in deep, core and REM sleep. The analysis puts your week, rhythm and average side by side.
The app already calculates these connections today. A connection is not a cause — but it shows you where to look.
The sleep detective checks caffeine in the 6 hours before falling asleep and whether you drank more than 400 ml in the last 90 minutes.
If you've logged a cortisol value, it appears on the cortisol card next to your rhythm.
In the Cycle area, you can note “slept badly” or “sleep disturbance” as a symptom.
Stress of 7 out of 10 or more on that day is one of the detective's findings. From ten shared days, your history shows stress and sleep together.
If you were outside clearly less than usual on a day, the detective checks it. From ten shared days, your history places sleep and sun side by side.
The detective checks intense training in the 3 hours before sleep. From ten shared days, your history shows training and sleep as a scatter chart.
If you ate at least 200 kcal in the 2 hours before sleep, the detective names it as a possible cause of a short night.
A low mood on that day is named by the detective as a possible cause — you log your mood in the Mindset area.
In a study of about 61,000 people (UK Biobank, movement sensors), a regular sleep-wake rhythm was linked to a 20 to 48 % lower risk of death. It predicted risk better than sleep duration. That is an association, not proof of a cause.1
In a review of studies, a warm shower or bath (40 to 42.5 °C) about one to two hours before bed shortened the time it took to fall asleep.2
Heat makes sleep more restless and shortens deep sleep and REM sleep. Air the room before bed and cover up so that you are neither hot nor cold.3
400 mg of caffeine taken six hours before bed measurably shortened sleep time in one study. Remember cola, black tea and green tea, too.4
Alcohol makes you sleepy faster but disturbs the second half of the night: sleep becomes more restless and the first REM sleep comes later.5
People who read on a light-emitting device in the evening fell asleep later, had less melatonin and were less alert the next morning than with a printed book.6
Nicotine lengthens the time to fall asleep and fragments sleep. Evening exercise does not harm sleep in studies; only very intense training that ends less than an hour before bedtime can affect falling asleep and sleep duration.78
A new article every Friday.
Sleep & Recovery · Sleep hygieneSleep Hygiene: 7 evidence-based rules for better sleepGood sleep is not a matter of luck but of habit. These seven rules are well supported by science – and most of them cost you nothing.Read more
Sleep · Nutrition & SleepCaffeine & sleep: how long caffeine really disrupts your sleepThe espresso after lunch sounds harmless. But caffeine has a half-life of five to seven hours — and still measurably disrupts your sleep at night, even if you feel awake in the evening.Read more
Sleep · EnvironmentSleep temperature: why sleeping cool is betterYour body needs to lower its core temperature in order to fall asleep. This makes room temperature one of the most powerful yet most underestimated levers for good sleep – with a clear recommendation from research.Read moreEvery statement comes with its source; the figures are taken from the linked studies and opinions (checked on 1 November 2026). This is context from the research, not a diagnosis and no substitute for medical advice.
Get YUTHI from the App Store — iPhone first, from November 2026.
