An overnight sleep study returns a report full of stages, indices and totals. Those numbers come from a human applying published rules to short windows of recorded signal.
Several signals are recorded at once
A standard study records brain activity, eye movement and muscle tone, alongside airflow, breathing effort, blood oxygen, heart rhythm and leg movement.
Stage determination depends on the first three, because stages are defined by patterns in brain waves combined with what the eyes and muscles are doing.
The remaining channels exist to detect breathing and movement events, which is why a study aimed at sleep apnea uses different equipment from one aimed at seizures.
Sleep is scored in thirty-second windows
The recording is divided into consecutive thirty-second segments, and each segment is assigned a single stage according to whichever pattern predominates within it.
That convention dates from the era of paper recordings moving at a fixed speed, and it survived because comparability across laboratories depends on everyone using the same window.
A consequence is that transitions shorter than the window disappear, so a report describes sleep at a resolution coarser than the underlying physiology.
Events are defined by thresholds
A breathing event is counted when airflow drops by a specified amount for a specified duration, sometimes with an additional requirement involving oxygen or arousal.
Those thresholds are conventions agreed by professional bodies, and they have been revised over time, which means indices from different eras are not directly comparable.
Two technologists scoring the same night generally agree closely on totals while disagreeing on individual segments, which is expected rather than a sign of error.
The laboratory night is not a typical night
Sleeping with electrodes attached in an unfamiliar room usually produces less sleep and more awakenings than the same person experiences at home.
Laboratories account for this in interpretation, and home testing exists partly to avoid it, though home devices record fewer channels and cannot determine stages the same way.
This is why a home test can identify a breathing problem while being unable to characterize the sleep architecture surrounding it.
What the report is for
The numbers exist to support a clinician's interpretation alongside history and examination, not to stand alone as a verdict on how well someone sleeps.
An index that looks alarming may be unremarkable in context, and a normal-looking study does not rule out a problem the test was not designed to detect.
Ordering, interpreting and acting on a sleep study is clinical work, and consumer trackers do not produce data of this kind despite using similar vocabulary.