Clinician-Focused Overview and Developments in Polysomnography

TL;DR

This clinician-focused narrative review summarizes the indications, technical components, and limitations of polysomnography (PSG), and situates it against home sleep apnea testing (HSAT) and emerging consumer/wearable and precision-medicine trends. PSG remains the gold-standard diagnostic test for suspected obstructive or central sleep apnea and is essential for nocturnal hypoventilation, periodic limb movements, central hypersomnias, and REM sleep behavior disorder, though cost, access, and night-to-night variability constrain its use.

Key findings

• PSG is required (not merely helpful) for several diagnoses: nocturnal hypoventilation disorders, periodic limb movements of sleep, central nervous system hypersomnia (in the absence of CSF hypocretin) when combined with the MSLT, and REM sleep behavior disorder (RBD) — the only parasomnia for which PSG is required, because documentation of REM sleep without atonia plus videographic or historical dream enactment is needed.
• RBD is described as a prodromal biomarker of neurodegenerative disease (Parkinson's disease, Lewy body dementia, multiple system atrophy), making PSG central to identifying at-risk individuals; video PSG with expanded EMG montage is an essential requirement for the diagnosis.
• HSAT is acceptable for high-probability OSA without comorbidities, with comparative studies showing similar diagnosis and CPAP treatment rates in patients without cardiovascular disease; PSG is recommended instead for congestive heart failure, neuromuscular weakness, suspected hypoventilation, chronic opioid use, suspected parasomnia or central sleep apnea, stroke history, and evaluation of central hypersomnolence or PLMS/RLS.
• A cited study of asymptomatic older adults found 75% met SRBD criteria by AHI > 15 alone, with no difference in subjective sleep quality, daytime sleepiness, or cognitive performance, but greater amyloid deposition and altered metabolism in posterior cingulate cortex and precuneus on MRI/PET; separate cited biomarker work found greater amyloid deposition in untreated OSA and lower serum/CSF amyloid versus controls and CPAP-treated patients.
• Scoring thresholds reported: STOP-BANG ≥ 3 indicates high OSA risk (0–24-point Epworth Sleepiness Scale described, higher = greater sleepiness); abnormal PLMS index is > 15/h in adults and > 5/h in children; PLMD requires PLMSI > 15/h in adults; narcolepsy/IH criteria include mean sleep latency ≤ 8 min and ≥ 2 SOREM on MSLT, with SOREM within 15 min on preceding PSG counting as one; the yield of capturing an epileptiform discharge on PSG is < 50%.
• Consumer wearables are a growing trend but few have received appropriate validation; only a few commercial devices (e.g., Dreem2) use EEG for staging, and early validation studies show feasible data acquisition with a moderate range of group-level agreement on blinded AASM-scored staging.

Methods

• Design: narrative, clinician-focused review summarizing indications for PSG, its limitations and challenges, HSAT options, and new developments and trends.
• Data/sample: no primary data; the paper synthesizes published literature, AASM practice parameters and the AASM Scoring Manual (version 2.6), ICSD-3 diagnostic criteria, and the 2017 USPSTF OSA screening recommendation, illustrated with example PSG tracings, hypnograms, and figures from the authors' practice.
• Analysis: descriptive synthesis organized by indication (excessive daytime sleepiness, sleep apnea, narcolepsy, movement disorders, parasomnias, nocturnal seizure), limitations, and trends; includes tables comparing standard versus additional PSG channels, STOP-BANG scoring, HSAT versus PSG indications, central hypersomnolence criteria, and HSAT versus PSG elements.

Limitations

• Stated by the paper: a single night may not reflect an individual's normal sleep due to night-to-night variability in many sleep disorders; the "first-night" effect can produce limited sleep time and worsened quality with less deep sleep; HSAT lacks a sleep technologist to correct technical issues mid-study, cannot be used with auto-PAP in CSA, hypoventilation syndromes, CHF, or COPD, and lacks EEG so cannot assess cortical arousals or RERAs, meaning it should not be used to "rule out" OSA; accuracy and quality of portable testing remain to be studied.
• Stated by the paper: PSG practice variations persist in technical and interpretation practices despite AASM standardization.
• Evident from the text (not framed as a limitation by the authors): this is a narrative review without a stated search strategy, inclusion/exclusion criteria, or systematic appraisal; several cited findings (e.g., the 75% SRBD figure and amyloid results) are drawn from single external studies cited in brief rather than synthesized quantitatively, and the review notes that a full discussion of MSLT/MWT is beyond its scope.