Polysomnography (Sleep Study): When to Order This Test

What Polysomnography Actually Measures

Polysomnography (PSG) is a supervised, in-laboratory sleep study that simultaneously records brain electrical activity (EEG), eye movements (EOG), chin and limb muscle tone (EMG), nasal and oral airflow, respiratory effort via chest and abdominal belts, pulse oximetry, heart rhythm (ECG), and body position. A trained technologist monitors the recording in real time, adjusting sensors and, when indicated, initiating a split-night CPAP titration if the apnea-hypopnea index exceeds 40 events per hour during the first two hours of recording [1].

The American Academy of Sleep Medicine (AASM) scoring manual defines an apnea as a 90% or greater reduction in airflow lasting at least 10 seconds, and a hypopnea as a 30% or greater airflow reduction accompanied by either a 3% oxygen desaturation or an EEG arousal [2]. These definitions matter because small changes in scoring criteria can shift a patient's AHI by several points, sometimes crossing a treatment threshold. The test generates between 800 and 1,000 pages of raw data per night, which a board-certified sleep medicine physician then interprets into a structured report. PSG remains the only test that can reliably distinguish obstructive from central apneas, stage narcolepsy-associated REM abnormalities, and quantify periodic limb movements during sleep [1].

Clinical Indications: When a Sleep Study Is Warranted

Order polysomnography when the clinical picture suggests sleep-disordered breathing that cannot be adequately evaluated with home testing alone, or when a non-respiratory sleep disorder is on the differential. The AASM clinical guideline published in the Journal of Clinical Sleep Medicine lists the following primary indications [3]:

Suspected obstructive sleep apnea with complicating factors. While home sleep apnea testing (HSAT) is appropriate for adults with a high pretest probability of moderate to severe OSA and no significant comorbidities, in-lab PSG is required when the patient has congestive heart failure, chronic opioid use, neuromuscular disease, or a history of stroke [3]. A negative or inconclusive HSAT in a patient with persistent symptoms also warrants follow-up PSG.

Central sleep apnea syndromes. HSAT devices do not differentiate central from obstructive events. Any clinical suspicion of Cheyne-Stokes respiration, treatment-emergent central apnea, or opioid-associated central apnea requires in-lab PSG [1].

Narcolepsy evaluation. The International Classification of Sleep Disorders, Third Edition (ICSD-3), requires a preceding overnight PSG before a multiple sleep latency test (MSLT) to rule out other causes of excessive daytime sleepiness [4].

Parasomnias with injury risk. REM sleep behavior disorder (RBD), where patients physically act out dreams, requires PSG to document REM without atonia [4]. This is not optional. RBD also serves as a prodromal marker for alpha-synuclein neurodegenerative disease, making the PSG finding clinically significant beyond the sleep complaint itself.

CPAP titration. When auto-titrating PAP therapy is insufficient (residual AHI above 10 on downloaded data, persistent symptoms, or complex sleep apnea), an attended in-lab titration study establishes the optimal fixed pressure [3].

OSA Prevalence and Why It Remains Underdiagnosed

Obstructive sleep apnea affects far more people than most clinicians realize. A 2019 analysis in The Lancet Respiratory Medicine estimated that 936 million adults aged 30 to 69 worldwide have mild to severe OSA (AHI of 5 or more), and 425 million have moderate to severe disease (AHI of 15 or more) [5]. In the United States, the Wisconsin Sleep Cohort Study found that 24% of men and 9% of women aged 30 to 60 met criteria for OSA on polysomnography, yet roughly 80% of moderate to severe cases remained undiagnosed at the time of the study [6].

The gap between prevalence and diagnosis persists for several reasons. Bed partners are often the first to notice apneas, and patients who sleep alone may attribute daytime fatigue to stress, aging, or poor sleep hygiene. Screening questionnaires like STOP-BANG have a sensitivity of 90% for moderate to severe OSA at a score of 3 or higher, but specificity is only about 49%, generating both false reassurance in low scorers and referral fatigue from high false-positive rates [7]. The AASM position statement recommends that clinicians screen for OSA using validated tools in patients with hypertension, type 2 diabetes, heart failure, atrial fibrillation, pulmonary hypertension, and treatment-resistant depression, as untreated OSA worsens outcomes in each of these conditions [3].

The Testosterone Connection: Why Sleep Studies Matter for Hormone Therapy

The Endocrine Society's 2018 clinical practice guideline for testosterone therapy in men with hypogonadism states: "We recommend against testosterone therapy in men with... untreated obstructive sleep apnea" and advises clinicians to "evaluate the patient for... sleep apnea... before and during testosterone treatment" [8]. This is not a minor footnote. It is a strong recommendation based on moderate-quality evidence.

Testosterone influences upper-airway collapsibility through several mechanisms. A randomized, placebo-controlled crossover trial of intramuscular testosterone (250 mg every 2 weeks for 6 weeks) in 17 obese men without baseline OSA found that testosterone administration increased the AHI from a median of 2.1 to 9.1 events per hour and reduced total sleep time spent above 90% oxygen saturation [9]. The effect reversed after washout. In men with pre-existing mild OSA, testosterone replacement may push AHI into the moderate or severe range, and the Endocrine Society guideline recommends repeating a sleep study at 3 to 6 months after initiating therapy if symptoms of sleep-disordered breathing develop [8].

Dr. Shalender Bhasin, principal investigator of the Testosterone Trials (TTrials) and professor of medicine at Brigham and Women's Hospital, has written: "Testosterone treatment of older men with low testosterone was associated with a small but significant increase in the apnea-hypopnea index" [10]. The TTrials enrolled 788 men aged 65 or older with serum testosterone below 275 ng/dL, and the testosterone-treated group showed a mean AHI increase of 2.5 events per hour above placebo at 12 months. While this magnitude may seem small, for a man whose baseline AHI is 13 (mild OSA, just below the moderate threshold), an increase of 2.5 events tips him into moderate disease.

The clinical takeaway is direct: any man starting testosterone replacement therapy (TRT) who scores 3 or higher on the STOP-BANG questionnaire, has a neck circumference above 17 inches, or reports witnessed apneas should undergo polysomnography or, at minimum, a home sleep apnea test before treatment initiation [8]. Men already on CPAP who begin TRT should have their device data reviewed at 3-month intervals to detect rising residual AHI.

How to Interpret Polysomnography Results

The AHI is the headline number, but it is not the only clinically relevant metric. A complete PSG report includes oxygen nadir, time spent below 90% SpO2 (T90), sleep architecture percentages, arousal index, periodic limb movement index (PLMI), and REM-specific AHI.

AHI thresholds. The AASM defines normal as fewer than 5 events per hour, mild OSA as 5 to 14, moderate as 15 to 29, and severe as 30 or more [2]. Insurance payers typically authorize CPAP at an AHI of 15 or greater, or at an AHI of 5 to 14 with documented symptoms (excessive daytime sleepiness, impaired cognition, mood disorder, hypertension, ischemic heart disease, or history of stroke) [3].

Oxygen burden. A patient may have a modest AHI of 12 but spend 15% of total sleep time below 90% saturation. This oxygen burden carries independent cardiovascular risk. The Sleep Heart Health Study, a community-based cohort of 6,441 adults followed for a median of 8.2 years, found that those spending more than 10% of sleep time below 90% SpO2 had 1.5 times the hazard of all-cause mortality compared to those with minimal desaturation, after adjusting for AHI [11].

REM-predominant OSA. Some patients have a normal overall AHI but an elevated REM AHI (often above 20 events per hour). REM-predominant OSA is more common in women, younger patients, and those with lower BMI. Because REM occupies only 20 to 25% of total sleep time, overall AHI can mask significant REM-specific disease. The clinical significance is debated, but observational data associate REM OSA with insulin resistance and hypertension independent of overall AHI [12].

Periodic limb movements. A PLMI above 15 per hour is considered elevated. Periodic limb movement disorder is diagnosed only when the PLMI is elevated, the patient reports sleep disruption or daytime impairment, and no other explanation (such as restless legs syndrome alone or medication effect) accounts for the movements [4].

Home Sleep Apnea Testing vs. In-Lab Polysomnography

Home sleep apnea testing uses a portable device that typically records nasal airflow, respiratory effort, and pulse oximetry without EEG. It is less expensive (often $300 to $600 vs. $1,500 to $3,000 for in-lab PSG) and more convenient, but it has limitations [3].

The AASM guideline specifies that HSAT is appropriate only for adults aged 18 or older with a high pretest probability of moderate to severe OSA and no significant cardiopulmonary comorbidity, no suspected non-respiratory sleep disorder, and no chronic opioid use [3]. HSAT underestimates AHI because it records total recording time rather than actual sleep time. A patient who lies awake for 2 of 8 recorded hours will have the denominator inflated, producing a falsely low AHI. The AASM recommends that a negative HSAT in a symptomatic patient should be followed by in-lab PSG rather than accepted as definitive [3].

The American Academy of Sleep Medicine's 2017 position paper further states: "HSAT is not appropriate for the diagnosis of OSA in patients with significant comorbid medical conditions... that may degrade the accuracy of HSAT" [13]. For patients being evaluated in the context of hormone therapy initiation, particularly those with obesity (BMI of 30 or above), metabolic syndrome, or known cardiovascular disease, in-lab PSG provides the higher diagnostic confidence that clinical decision-making demands.

Ordering Logistics: Referral, Prior Authorization, and Preparation

Most insurance plans require a referral from a primary care provider or specialist to a board-certified sleep medicine physician, along with documented clinical indication, before authorizing polysomnography. Common documentation requirements include:

A completed sleep-specific questionnaire such as the Epworth Sleepiness Scale (ESS) score. An ESS of 10 or above suggests excessive daytime sleepiness [14]. A STOP-BANG score of 3 or more supporting moderate to high pretest probability. Relevant comorbidity documentation (hypertension, atrial fibrillation, type 2 diabetes, planned TRT).

Patients should be instructed to avoid alcohol for 24 hours and caffeine after noon on the day of the study. Sedative-hypnotics and benzodiazepines should be discussed with the ordering physician, as some protocols require a 2-week washout to avoid suppression of arousals that would falsely lower AHI, while others deliberately test the patient on their current medication regimen [1].

Expect results within 7 to 14 business days. The interpreting sleep physician scores the study, assigns an AHI and diagnostic classification, and sends recommendations back to the referring clinician. If the AHI exceeds 15 events per hour and the patient is being considered for TRT, the standard of care per Endocrine Society guidelines is to treat the OSA (typically with CPAP or mandibular advancement device) and reassess symptoms of hypogonadism before starting exogenous testosterone, because OSA itself suppresses gonadotropin-releasing hormone pulsatility and contributes to secondary hypogonadism [8].

Metabolic Consequences of Untreated OSA

Untreated obstructive sleep apnea is not just a nuisance. It is a metabolic disease. The intermittent hypoxia and sleep fragmentation caused by recurrent apneas activate sympathetic nervous system overdrive, increase oxidative stress, and promote systemic inflammation.

A meta-analysis of 16 prospective studies published in The Lancet Respiratory Medicine found that moderate to severe OSA (AHI of 15 or more) was associated with a 63% increased risk of incident type 2 diabetes after adjustment for BMI and other confounders (pooled relative risk 1.63, 95% CI 1.09 to 2.45) [15]. The relationship appears bidirectional. Insulin resistance promotes adipose deposition in the upper airway and tongue, worsening collapsibility.

For patients on GLP-1 receptor agonists, weight loss may improve OSA severity. The SURMOUNT-OSA trial (N=469) evaluated tirzepatide 10 mg or 15 mg vs. placebo in adults with moderate to severe OSA and obesity. At 52 weeks, tirzepatide reduced AHI by approximately 30 events per hour from a baseline mean of roughly 51.5 events per hour, compared with a reduction of about 6 events per hour in the placebo group [16]. Dr. Atul Malhotra, chief of pulmonary, critical care, and sleep medicine at UC San Diego and co-investigator of SURMOUNT-OSA, stated: "The magnitude of AHI reduction with tirzepatide was remarkable and suggests that weight loss may be a viable primary therapy for many patients with obesity-related OSA" [16].

These findings do not eliminate the need for polysomnography. They reinforce it. Patients need a baseline AHI to measure treatment response, whether the intervention is CPAP, weight loss via GLP-1 therapy, or both.

When to Repeat a Sleep Study

A follow-up polysomnography is indicated in specific clinical scenarios. After significant weight change (gain or loss of 10% or more of body weight), the AHI may shift enough to change treatment decisions [3]. If a patient on CPAP has persistent daytime sleepiness despite good adherence (defined as 4 or more hours per night on 70% or more of nights), a repeat PSG can identify residual events, treatment-emergent central apnea, or a concurrent sleep disorder. After surgical intervention for OSA, such as uvulopalatopharyngoplasty or maxillomandibular advancement, a post-operative PSG at 3 to 6 months documents treatment success or failure. And as noted, initiation of testosterone therapy in a man with borderline or mild OSA warrants reassessment at 3 to 6 months per Endocrine Society recommendations [8].

A patient whose AHI drops below 5 events per hour on a follow-up study performed off CPAP after major weight loss (such as that achieved with bariatric surgery or sustained GLP-1 therapy) may be a candidate for a supervised CPAP discontinuation trial, though long-term recurrence rates remain poorly characterized [3].