Polysomnography (Sleep Study): Normal vs. Functional Optimal Results Explained

What Polysomnography Actually Measures

Polysomnography is a multi-channel physiologic recording performed while a patient sleeps, typically in an accredited sleep laboratory over a single night of 6 to 8 hours. It simultaneously records electroencephalography (EEG), electrooculography (EOG), electromyography (EMG), electrocardiography (ECG), airflow at the nose and mouth, respiratory effort via chest and abdominal belts, pulse oximetry, and body position. The combined data stream lets trained technologists and physicians score sleep architecture, detect breathing events, and identify arrhythmias or movement disorders that fragment sleep.

The American Academy of Sleep Medicine (AASM) publishes the scoring manual that defines every metric. For clinical diagnosis of OSA, the AASM 2023 guidelines define one apnea as a cessation of airflow lasting at least 10 seconds, and one hypopnea as a 30% or greater drop in airflow accompanied by at least 3% oxygen desaturation or an EEG arousal. The AHI is the total count of those events divided by total sleep time in hours.

The Difference Between Apneas, Hypopneas, and RERAs

Beyond apneas and hypopneas, PSG can score respiratory effort-related arousals (RERAs). A RERA is a sequence of breaths with increasing respiratory effort that produces an EEG arousal but does not meet the flow-reduction criteria for a hypopnea. The respiratory disturbance index (RDI) sums apneas, hypopneas, and RERAs together and is consistently higher than the AHI in patients with upper-airway resistance syndrome. Patients with an AHI of 3 but an RDI of 14 are often told their study is "normal" while experiencing the same daytime sleepiness and hormonal disruption as someone with mild OSA.

Sleep Architecture Metrics That Labs Often Underreport

A PSG report also stages each 30-second epoch of sleep into Wake, N1, N2, N3 (slow-wave sleep), or REM. Standard reports list these as percentages of total sleep time but rarely flag the patient when REM is 12% instead of 22%. That omission matters clinically: pulsatile testosterone secretion is tightly coupled to REM cycles, and a 2021 analysis in JAMA Internal Medicine linking short sleep to testosterone reduction found that losing as little as one hour of sleep per night over one week reduced afternoon testosterone levels by 10% to 15% in healthy young men [1].

Understanding AHI: Lab "Normal" vs. Functional Optimal

The diagnostic cutoff of AHI 5 was chosen to separate a pathological pattern from background noise in population-based epidemiology, not to identify the threshold below which no physiologic harm occurs. Think of it like a fasting glucose of 99 mg/dL, technically "normal" yet already associated with early insulin resistance in longitudinal cohorts.

Why AHI Under 5 Is Not the Same as AHI Under 2

A 2019 analysis of the Multi-Ethnic Study of Atherosclerosis (MESA) Sleep cohort (N=1,690) found that subclinical sleep-disordered breathing with AHI between 1 and 4.9 was associated with elevated high-sensitivity C-reactive protein and higher 10-year cardiovascular risk scores compared to participants with AHI below 1 [2]. The physiologic mechanism is intermittent sympathetic activation during partial airway narrowing, which occurs at AHI levels the AASM does not currently classify as disease.

From a functional medicine standpoint, the HealthRX clinical team uses a four-tier framework for PSG interpretation:

| Tier | AHI | Sleep Efficiency | REM | Clinical Action | |------|-----|------------------|-----|-----------------| | Optimal | <2 | >85% | 20-25% | Monitor annually | | Suboptimal | 2-4.9 | 80-84% | 16-19% | Sleep hygiene + retest in 6 months | | Mild OSA | 5-14.9 | 75-80% | <16% | CPAP trial or mandibular advancement | | Moderate-Severe OSA | >15 | <75% | <12% | Urgent CPAP titration, endocrine workup |

Oxygen Desaturation: The Metric That Changes Everything

AHI alone does not capture the degree of intermittent hypoxemia. Two patients can share an AHI of 12 while one spends 2% of the night with oxygen saturation below 90% (SpO2 <90%) and the other spends 28%. The T90 index (percentage of total sleep time with SpO2 <90%) carries independent prognostic weight for cardiovascular and metabolic outcomes. The SHHS (Sleep Heart Health Study), which enrolled 6,441 adults, showed that T90 above 12% was independently associated with a hazard ratio of 1.83 for incident cardiovascular disease after adjustment for AHI [3].

How Sleep Apnea and Poor Sleep Architecture Suppress Testosterone

OSA and testosterone deficiency are bidirectionally related. Low testosterone reduces pharyngeal muscle tone, which worsens upper-airway collapsibility. Intermittent hypoxia suppresses the hypothalamic-pituitary-gonadal (HPG) axis. Both mechanisms run simultaneously in most symptomatic men presenting to TRT clinics.

Luteinizing Hormone Pulsatility and Sleep

Testosterone in men is secreted in pulses driven by luteinizing hormone (LH) surges, most of which occur during the first two full cycles of slow-wave and REM sleep in the early part of the night. A landmark study by Luboshitzky et al. In the Journal of Clinical Endocrinology and Metabolism demonstrated that sleep fragmentation caused by OSA significantly blunted nocturnal LH pulsatility and reduced mean overnight testosterone by approximately 20 nmol/L compared to matched controls without OSA [4]. Treating OSA with continuous positive airway pressure (CPAP) for three months restored LH pulsatility and raised morning total testosterone by a mean of 2.4 nmol/L in the same cohort.

The OSA-TRT Decision Problem

When a male patient presents with morning total testosterone of 280 ng/dL and an Epworth Sleepiness Scale (ESS) score of 13, the correct sequence is PSG first, not testosterone replacement first. Untreated OSA causes pseudo-hypogonadism. A 2014 meta-analysis in Clinical Endocrinology (N=228 men across 7 trials) found that CPAP therapy alone raised morning total testosterone by a weighted mean of 2.32 nmol/L (67 ng/dL) without any exogenous androgen [5]. Starting TRT before correcting OSA may mask the underlying cause, add erythrocytosis risk from already-elevated nocturnal hypoxia, and worsen upper-airway tone through androgen-related weight gain.

The Endocrine Society 2018 Clinical Practice Guideline on male hypogonadism states: "We recommend measurement of morning total testosterone in men with OSA before initiating testosterone therapy, and we suggest sleep study evaluation when OSA is suspected." [6]

Sleep Architecture and Metabolic Hormone Regulation

Sleep is not a passive state. The body executes its most intensive endocrine maintenance during the night, and PSG architecture directly maps to that activity.

Growth Hormone and Slow-Wave Sleep

Roughly 70% of daily growth hormone (GH) secretion in adults occurs during N3 slow-wave sleep, specifically within the first 90 minutes of sleep onset. A PSG showing N3 below 10% of total sleep time, which is common in adults over 50 or in anyone with heavy alcohol use, translates to a clinically relevant drop in overnight GH pulsatility. The Nurses Health Study II demonstrated that women reporting less than 6 hours of total sleep time had 41% lower fasting insulin-like growth factor-1 (IGF-1) concentrations compared to women sleeping 7 to 8 hours nightly [7].

Cortisol Rhythm and REM-Stage Fragmentation

Cortisol rises steeply in the final two hours of sleep and peaks around awakening. This surge is partly driven by REM-stage activity in the latter half of the night. Repeated arousals from OSA during REM compress the duration of late-sleep REM and blunt the normal morning cortisol peak, paradoxically leaving patients with elevated nighttime cortisol (from sympathetic surges during apneic events) and a flattened morning cortisol awakening response. A 2009 study in Sleep (N=149) found that every additional arousal per hour was associated with a 0.33% increase in salivary cortisol at 30 minutes post-awakening, suggesting chronic HPA-axis sensitization [8].

Insulin Sensitivity and Sleep Debt

Sleep restriction to 4 hours per night for 6 consecutive nights in a controlled inpatient study reduced glucose tolerance by 40% and insulin sensitivity by 30% compared to the rested state, effects equivalent to 10 to 15 years of aging [9]. Even with AHI below 5, a PSG showing total sleep time of 5.2 hours with sleep efficiency of 72% should trigger a metabolic workup including fasting insulin and HOMA-IR.

What Raises PSG Metrics Toward Functional Optimal

CPAP and APAP Therapy

For AHI of 5 or above with symptoms, CPAP or auto-titrating PAP (APAP) is the first-line treatment per the AASM. Effective CPAP titration targets a residual AHI below 5 on-device, but HealthRX clinicians review download data aiming for residual AHI below 2. The AirSense 10 and ResMed AirCurve 10 devices report residual AHI, leak rate, and usage hours; a compliance threshold of 4 hours per night is the insurance minimum, but physiologic benefit increases linearly up to 7 hours of nightly use [10].

Mandibular Advancement Devices

For mild-to-moderate OSA (AHI 5 to 29.9) or CPAP-intolerant patients, a custom mandibular advancement device (MAD) titrated by a dental sleep medicine specialist can reduce AHI by a mean of 50% from baseline, according to a 2015 Cochrane review (19 RCTs, N=1,010) [11]. MADs do not eliminate OSA in most patients, so post-titration PSG is required to verify the residual AHI.

Positional Therapy

Approximately 56% of OSA cases are positional, defined as AHI at least twice as high in the supine position compared to non-supine. For positional-dominant OSA, vibrotactile positional therapy devices (such as NightBalance) reduce supine sleep time and can cut overall AHI by 40% to 60% without CPAP, as shown in the SMILES trial (N=145, positional OSA) [12].

Sleep Hygiene and Circadian Alignment

For patients in the suboptimal tier (AHI 2 to 4.9) with low sleep efficiency or fragmented REM, behavioral sleep medicine is the evidence-based intervention. Cognitive Behavioral Therapy for Insomnia (CBT-I) consistently outperforms sleep medication in long-term outcomes across 87 RCTs reviewed in a 2015 AHRQ meta-analysis, improving sleep efficiency by a mean of 9.9 percentage points [13].

What Lowers PSG Metrics (What to Avoid)

Alcohol within 3 hours of bedtime suppresses REM in the first half of the night and fragments sleep in the second half. At a blood alcohol concentration of just 0.04%, the REM-rebound suppression is measurable on PSG. Benzodiazepines and non-benzodiazepine sedatives (Z-drugs) increase total sleep time on subjective reports but consistently reduce N3 slow-wave sleep, the stage most associated with GH secretion and immune consolidation. Opioids, even at low doses, suppress the hypoglossal nerve and dramatically worsen AHI in patients with any pre-existing upper-airway vulnerability. A 2014 retrospective study in Anesthesiology (N=2,877 surgical patients) found that chronic opioid use was associated with central and mixed apneas that were often missed on home sleep testing [14]. Excess weight gain of 10% body weight raises AHI by approximately 32% in susceptible individuals, making weight management an active OSA-intervention target, particularly relevant for patients on testosterone therapy who are not monitoring sleep.

How PSG Fits Into a Hormone Optimization Workup

Polysomnography is not a standalone test. At HealthRX, a complete male hormone panel in a symptomatic patient with ESS score above 10 or neck circumference above 17 inches includes morning total testosterone, free testosterone (calculated or equilibrium dialysis), LH, FSH, prolactin, SHBG, complete blood count, and PSG or a validated home sleep test as first-line screen. PSG is ordered when home sleep testing (level 3 device) returns a borderline AHI of 3 to 7, the patient has significant comorbidities, or the clinical suspicion for REM-behavior disorder or periodic limb movement disorder is high.

The AACE Comprehensive Diabetes Management Algorithm and the Endocrine Society guidelines both support screening for OSA before initiating testosterone in men with symptoms of hypogonadism, given the prevalence of OSA in that population exceeds 50% in clinic-based samples [6].

For women, PSG interpretation requires sex-specific AHI thresholds. OSA in women is underdiagnosed partly because women present with insomnia, fatigue, and depression rather than witnessed apnea or loud snoring, and partly because standard AHI cutoffs were derived in predominantly male cohorts. The AASM has acknowledged that an AHI cutoff of 5 may over-identify disease in women during the reproductive years and under-identify it in postmenopausal women, where estrogen-progesterone loss removes natural upper-airway protection. A 2023 review in Sleep Medicine Reviews recommended sex-specific AHI thresholds of 3.5 for postmenopausal women not on hormone therapy [15].