Why does teeth grinding change so much from night to night?
The maker of Hushbean · Updated
The short answer
Sleep grinding varies a great deal between nights even when nothing has changed. Across 25 nights of one person, recorded with the same phone in the same room, grinding ranged from 1.4 to 13 clips an hour, a ninefold spread. Research finds the same pattern, so one or two nights tell you very little about a habit or a treatment.
25 nights, measured
These are the nights of the person who makes Hushbean, from 2 September to 3 November 2026: the same iPhone, the same bedside table, the same bedroom, and the same version of the detector for every night. Each bar is the number of grinding clips found per hour of sleep.
- grinding clips per hour, quietest night to busiest
- 1.4 to 13.0
- grinding clips per hour, quietest night to busiest
- coefficient of variation: the typical night sat about half the average away from it
- 0.48
- coefficient of variation: the typical night sat about half the average away from it
- of sleep across the 25 nights
- 181 hours
- of sleep across the 25 nights
How these were counted: nights of three hours or more; the first twenty minutes and the last two left out; only clips that passed the app's own check and weren't rejected by ear. In a blind check of that check, 97% of the clips it showed were heard as grinding.
Research finds the same
In people followed over several sleep-lab nights, the number of grinding-noise episodes varied by about 54% from night to night within the same person (Lavigne et al., 2001). Some people who clearly grind fall below the diagnostic threshold on some nights for no visible reason. Sleep grinding is tied to brief arousals from sleep, and how often those happen changes nightly.
What this means if you're trying to change it
- One good night proves nothing. Nor does one bad night. A swing of this size happens with nothing changed.
- Compare blocks of nights. Several nights with a change against several without, recorded the same way.
- Give it weeks. Hushbean will not say anything about a habit until it has twelve comparable nights, and runs a change as a two-week try, alternating nights, so ordinary variation doesn't pass for an effect.
- Keep the recording the same. Moving the phone changes the numbers more than most habits do.
If you want to test something specific, such as alcohol, caffeine or stress, this guide explains what the research has found, and how to test it on your own nights.
Questions people ask
- Why did I grind more last night?
- Possibly nothing in particular. Grinding varies widely between nights in the same person with nothing changed. Look for a pattern across a couple of weeks before blaming one thing.
- How many nights does it take to see a pattern?
- More than most people expect: at least a week to know whether you grind, and several weeks to compare a habit. Hushbean waits for twelve comparable nights before it compares habits.
Read next
Sources
- Lavigne GJ, et al. Variability in sleep bruxism activity over time. J Sleep Res, 2001.
- Lavigne GJ, et al. Bruxism physiology and pathology: an overview for clinicians. Crit Rev Oral Biol Med, 2003.
- Hushbean's own blind check: 84 clips from 26 nights, heard by the person who slept through them, 1 November 2026.
About the author. Written by the person who makes Hushbean, who grinds their own teeth and has recorded more than 180 hours of their own sleep with it. They are not a dentist or doctor, and this page was not reviewed by one. General information, not medical advice.