When Your Sleep Score Becomes the Stressor

This piece looks at whether anxious preoccupation with sleep tracker scores, the pattern now called orthosomnia, has actually been measured, and what a general-population study found about how often it shows up. It does not cover treatment approaches for sleep-related anxiety or rank specific device brands.

Orthosomnia, the anxious fixation on a sleep tracker's score rather than how rested someone actually feels, has been formally measured in a general population sample rather than just described anecdotally. Depending on how strictly it was defined, a small to moderate share of regular tracker users in that sample met the criteria. That resolves the basic question of whether this is a documented pattern or just forum talk, at least for the population the study looked at.

When the Score Feels Like the Point

People describe two very different relationships with the same device. Some say checking their sleep score is genuinely calming, a way to feel like something is being tracked and managed. Others describe the opposite: lying awake doing math about what tonight's score will look like, or feeling deflated when a number doesn't match how good they thought last night felt.

That split is exactly what a cross-sectional study set out to quantify. Using a combination of an anxiety scale, an insomnia scale, and a scale built specifically to measure preoccupation with sleep, the team behind Jahrami et al. built a layered definition of orthosomnia and applied it to a general population sample. It wasn't a rare curiosity. Depending on how conservative the criteria were set, the identified rate ranged from a small single-digit share up to a more sizable double-digit share of people who regularly used sleep-tracking devices.

Is a Dropping Score About You, or About the Algorithm?

One recurring question doesn't involve anxiety directly but feeds it: why does a score drop when the total hours logged look basically unchanged? Part of the answer sits in how these devices actually build a sleep score in the first place, a separate question covered in how wearables actually track sleep stages.

Work on the modeling side gives a partial answer. Lee et al. tested a neural network that combined topological features of motion and heart rate data with circadian modeling, comparing its output against polysomnography in people wearing an Apple Watch. Adding those circadian and heart-rate-derived features improved the classification of REM and NREM sleep compared with a model using only raw wearable data, with most of the gain traced to the heart rate features. That's a reminder that stage and score calculations come from layered, evolving models rather than one fixed reading, and comparing that against how REM sleep actually shows up on a tracker makes the gap between raw data and a final number a little more visible.

None of that tells anyone whether their own score change is real or just an algorithm quirk. Hard to know from the outside. But that doesn't mean there's one universal answer to the question forums keep asking, whether a lower score is 'real' or just recalibration. Comparing how different brands classify sleep in the first place is its own separate topic, covered in how accurate Fitbit, Garmin, and WHOOP sleep tracking actually is.

What This Research Doesn't Settle

The orthosomnia study is useful precisely because it puts a number on something that had mostly been described in passing, a number from one cross-sectional sample only answers a narrower question than it might seem to.

It tells us how many people in that particular sample met a defined set of criteria at one point in time. What it doesn't tell us is whether using a tracker caused the anxious preoccupation, whether pre-existing anxiety about sleep led certain people to start tracking in the first place, or how the pattern behaves over months or years of use, at least not from this one snapshot.

Common questions