Basic Neurosciences · Paper A
Sleep architecture
Normal sleep cycles through distinct stages roughly every 90 minutes, and depression disturbs this pattern in specific, testable ways. For example, imagine an EEG recording from a patient with severe depression, it typically shows REM sleep starting earlier in the night than normal (reduced REM latency), a finding used as a research marker rather than a stand-alone diagnostic test. In practice, interpret sleep architecture as a pattern, not proof of one disorder.
REM sleep shows low-voltage fast EEG activity with muscle atonia (paralysis) and dreaming. Slow-wave sleep (stage N3) shows high-amplitude delta waves and dominates earlier in the night, while REM periods lengthen progressively toward morning. For example, imagine waking someone from the first third of the night versus near dawn: deep N3 sleep is more likely early, while a long REM period is more likely late. In practice, this means the proportion and timing of REM versus slow-wave sleep changes across a single night, so a sleep study's timing matters as much as what it measures.
Imagine two people who both sleep exactly eight hours, yet one wakes up refreshed and the other exhausted. Total sleep TIME tells you almost nothing about that difference; what matters is the internal STRUCTURE of the night, which stages happened, in what order, and how much of each. That structure is sleep architecture, and several major psychiatric disorders disturb it in specific, testable ways.
Sleep is not a single uniform state; it cycles through distinct, measurable stages across the night, and this pattern is called sleep architecture. The stages are conventionally divided into non-rapid eye movement (NREM) sleep, which itself has lighter and deeper substages, and rapid eye movement (REM) sleep, the stage most associated with vivid dreaming.
These architectural changes are used mainly as research markers of a disorder's underlying neurobiology rather than as a stand-alone diagnostic test. Diagnosis of a depressive episode, for example, still rests on clinical criteria, such as those in ICD-11 or DSM-5, not on a sleep study result in isolation. It's tempting to treat a sleep-stage abnormality as "proof" of a diagnosis; it is evidence FOR a pattern, not a substitute for clinical assessment.
NREM sleep progresses through stages, with stage N3, also called slow-wave sleep, representing the deepest stage, characterised on EEG (electroencephalogram, a recording of the brain's electrical activity) by high-amplitude, slow delta waves. Slow-wave sleep is thought to be the most physically restorative sleep stage and dominates earlier in the night. REM sleep, by contrast, shows low-voltage, fast EEG activity resembling wakefulness, paired with muscle atonia, meaning near-total paralysis of the skeletal muscles, which is thought to prevent a sleeper from physically acting out dreams.
Imagine waking a healthy volunteer from sleep in the first third of the night versus close to morning: the first waking is more likely to catch them in deep slow-wave sleep, while the second is more likely to catch them in a long REM period, since REM periods lengthen and NREM depth lightens progressively across a normal night's roughly 90-minute cycles.
Depression is classically associated with two specific architectural changes: shortened REM latency, meaning REM sleep begins earlier after sleep onset than the normal roughly 90 minutes, and reduced slow-wave sleep, meaning less of the deep, restorative N3 stage across the night. Consider a polysomnography (an overnight, multi-channel sleep study that is the standard investigation for characterising sleep architecture) recording from a patient with severe depression: it typically shows REM onset under 60 minutes rather than the normal 90, alongside a diminished proportion of slow-wave sleep. Most antidepressants, particularly SSRIs and TCAs, then SUPPRESS REM sleep further on treatment, delaying its onset and reducing its overall proportion across the night; clinicians should not mistake a treatment-induced change in REM sleep for a sign that the underlying illness is worsening.
Depression is the classic example taught for exams, but disturbed sleep architecture is not unique to it. Imagine comparing three patients referred for a sleep study: one with major depression showing shortened REM latency, one with obstructive sleep apnoea showing fragmented sleep with repeated brief arousals from oxygen desaturation, and one with narcolepsy showing REM sleep intruding abnormally early, sometimes even at sleep onset, alongside daytime sleep attacks and cataplexy (sudden loss of muscle tone triggered by strong emotion). Distinguishing these patterns is exactly why polysomnography, rather than history alone, is used as the definitive investigation when a primary sleep disorder is suspected rather than a mood disorder with secondary sleep disturbance.
Polysomnography remains the investigation of choice when sleep architecture genuinely needs to be characterised, for example in suspected primary sleep disorders or research settings, but it is not part of routine first-line assessment for straightforward depression, where clinical assessment and history-taking remain central. Screening for sleep disturbance as a SYMPTOM, rather than an EEG-based sleep study, is what most clinical encounters actually involve; polysomnography is reserved for cases where the pattern of disturbance itself (loud snoring with witnessed apnoeas, daytime sleep attacks, cataplexy) points toward a distinct primary sleep disorder needing its own diagnosis.
Management differs sharply depending on the underlying diagnosis, which is exactly why an accurate diagnosis, not just a description of poor sleep, must come first. Obstructive sleep apnoea is managed with continuous positive airway pressure (CPAP) and weight-related lifestyle measures rather than an antidepressant. Narcolepsy is managed with specialist referral and stimulant or REM-suppressing medication under specialist care, not standard psychiatric treatment. Where disturbed sleep architecture is secondary to depression itself, management is management of the depressive episode, following standard NICE-based pathways for depression, since correcting the underlying mood disorder typically normalises REM latency and slow-wave sleep over time rather than requiring a separate sleep-focused treatment.
Know the DIRECTION of each change rather than exact figures: depression SHORTENS REM latency and REDUCES slow-wave sleep, while effective antidepressant treatment further SUPPRESSES REM. If a stem instead describes REM intruding at sleep onset with cataplexy, that's narcolepsy, not depression, a common distractor pairing.
Normal sleep cycles through NREM (including slow-wave sleep) and REM roughly every 90 minutes; depression shortens REM latency and reduces slow-wave sleep, antidepressants suppress REM further, and polysomnography, not a sleep-stage finding alone, is what confirms a distinct primary sleep disorder when one is suspected.