Nightmares & Disorders
What a sleep study involves, and when it is worth having one
Twenty-odd sensors, a night in an unfamiliar bed, and a technician watching. Home testing has replaced it for many cases — but not for the ones where it matters most.
Polysomnography is the reference standard for sleep measurement, and having spent six years attaching people to it, I can say the most common question is whether anyone can sleep like that. Most people do, and the second night in a laboratory is usually better than the first — the first-night effect is well documented and is why some protocols discard it.
What gets measured
EEG — electrodes on the scalp, recording brain electrical activity. This is what determines sleep stage. Without EEG you cannot stage sleep, which is the single most important limitation of consumer devices.
EOG — electrodes beside the eyes, recording eye movement. Identifies REM.
EMG — electrodes on the chin, recording muscle tone. Chin atonia is the third criterion for REM. Additional leads on the legs detect periodic limb movements.
ECG — heart rhythm.
Airflow — a thermistor and a nasal pressure transducer at the nostrils.
Respiratory effort — belts around chest and abdomen. The distinction between obstructive and central apnoea depends on this: in obstructive events, effort continues against a closed airway; in central events, effort stops.
Pulse oximetry — oxygen saturation.
Position sensor, snore microphone, and video.
Home testing is appropriate for uncomplicated suspected obstructive sleep apnoea in someone with a high pre-test probability and no significant comorbidity.
Full laboratory study is needed for: suspected central sleep apnoea · significant cardiac, pulmonary or neuromuscular disease · suspected parasomnias, including REM sleep behaviour disorder · suspected narcolepsy · nocturnal seizures · a negative home study where suspicion remains high. Home tests cannot stage sleep and can therefore underestimate severity, because they divide events by time in bed rather than time asleep.
The multiple sleep latency test
Performed the day after an overnight study, to measure objective sleepiness. Five nap opportunities at two-hourly intervals, in a darkened room, with the patient instructed to try to sleep.
Two numbers matter: mean sleep latency across the naps, and the number of sleep-onset REM periods. A mean latency of eight minutes or less with two or more sleep-onset REM periods supports narcolepsy.
The test is easily invalidated. It requires adequate sleep in the preceding nights — documented, usually with a sleep diary and actigraphy — and withdrawal of REM-suppressing medications, especially antidepressants, for a period beforehand. Failing to do this produces false results in both directions and is a common reason for having to repeat it.
The maintenance of wakefulness test is the mirror image, measuring ability to stay awake, and is used for occupational assessments — pilots, drivers — rather than for diagnosis.
What actigraphy is for
A wrist-worn accelerometer worn for one to two weeks, estimating sleep and wake from movement.
It cannot stage sleep and it is imprecise about sleep onset. What it does well is establish patterns over time — which makes it the appropriate tool for circadian rhythm disorders, for suspected insufficient sleep, and for validating a sleep diary before an MSLT.
Consumer wearables use the same principle with less validation. They are reasonable for tracking your own trends and are not diagnostic instruments.
When to ask for a study
Suspected sleep apnoea — witnessed pauses, loud snoring with gasping, unrefreshing sleep despite adequate hours, morning headache, resistant hypertension, unexplained atrial fibrillation.
Excessive daytime sleepiness that is not explained by insufficient sleep. Falling asleep unintentionally, particularly while driving, is urgent.
Dream enactment — physically acting out dreams, especially in someone over fifty. This warrants a proper study, because REM sleep behaviour disorder is strongly associated with later Parkinson's disease and related conditions, and confirming it changes what happens next.
Unusual nocturnal behaviour where the distinction between parasomnia and nocturnal seizure matters.
Suspected narcolepsy — sleepiness with cataplexy, sleep paralysis or hypnagogic hallucinations.
When a study is not the answer
Chronic insomnia does not require polysomnography in most cases. It is a clinical diagnosis, and the treatment is CBT-I. A study is indicated only if another disorder is suspected underneath — apnoea, periodic limb movements — or if treatment has failed unexpectedly.
This matters because people with insomnia frequently want a study, on the reasonable assumption that measurement will explain something. What it usually shows is more sleep than they experienced, which is informative about sleep state misperception and does not change management.
Practical notes if you are having one
Keep your normal schedule beforehand. Do not nap that day, and do not deliberately sleep-deprive yourself hoping to sleep better in the laboratory — it invalidates the study. Avoid caffeine after midday and alcohol entirely. Wash your hair and skip conditioner and product; electrodes do not stick to it. Bring your own pillow if you want. Take your usual medications unless specifically told otherwise, and tell the technician exactly what you take.
And expect to sleep worse than usual. Everyone does, the technicians know, and the study is still interpretable.
