Polysomnography (Sleep Study) Results: How to Read Your Numbers and What Changes Over Time

What a Sleep Study Actually Measures

A polysomnography records at least seven data streams simultaneously: brain waves (EEG), eye movements (EOG), muscle activity (EMG), heart rhythm (ECG), airflow, respiratory effort, and blood oxygen (SpO2). The result is not one single number but a profile of how your body cycles through sleep stages, whether your airway collapses, and whether your brain and heart pay a price for it.

The American Academy of Sleep Medicine (AASM) defines polysomnography as the gold-standard diagnostic test for obstructive sleep apnea (OSA), central sleep apnea, parasomnias, and periodic limb movement disorder. The core metrics most clinicians focus on are the Apnea-Hypopnea Index (AHI), the Oxygen Desaturation Index (ODI), sleep efficiency, sleep stage percentages, and the respiratory disturbance index (RDI).

The Apnea-Hypopnea Index (AHI)

The AHI counts the average number of breathing pauses (apneas) or partial collapses (hypopneas) per hour of sleep. AASM severity thresholds are:

| AHI | Severity | |---|---| | <5 | Normal | | 5-14.9 | Mild OSA | | 15-29.9 | Moderate OSA | | ≥30 | Severe OSA |

Here is the catch for women: the standard AHI cutoff was derived largely from male populations. A 2019 analysis in Sleep Medicine found that women with moderate-to-severe symptoms frequently had AHIs in the mild range (5-15) because their events tend to be shorter, more hypopnea-dominant, and more clustered in REM sleep.

Oxygen Saturation Nadir and ODI

The SpO2 nadir is the lowest oxygen level recorded across the night. An SpO2 nadir above 90% is considered acceptable in most adults. The ODI counts desaturation events (drops of 3% or 4% below baseline, depending on which scoring rule is used) per hour. Clinical guidelines from the AASM recommend treatment when ODI is elevated even if the AHI alone does not meet the threshold for OSA, particularly in the presence of symptoms.

Sleep Architecture

Your night is divided into cycles of roughly 90 minutes, each containing light sleep (N1 and N2), deep slow-wave sleep (N3), and REM. Normative data published in Sleep across 2,400 adults show:

• N1: 5-10% of total sleep time
• N2: 45-55%
• N3: 13-23% (highest in young adults, declining with age)
• REM: 20-25%
• Sleep efficiency: >85%

N3 sleep is where growth hormone pulses occur, cellular repair happens, and memory consolidation is strongest. Losing N3 is not just a matter of feeling groggy. It carries metabolic consequences that matter specifically for women with PCOS, insulin resistance, and perimenopausal metabolic shifts.

Rate-of-Change Interpretation: What Your Results Should Look Like Over Time

Most women get one sleep study and are told they passed or failed. Rate-of-change interpretation means asking a different question: is your sleep quality stable, improving, or drifting in a direction that predicts future disease?

A landmark longitudinal analysis from the Wisconsin Sleep Cohort followed participants for up to 14 years and found that AHI increases by an average of 0.4 events per hour per year in the general population, but women accelerate sharply through the menopausal transition, sometimes gaining 5 to 10 AHI points within a few years.

The WomanRx rate-of-change framework for polysomnography uses three zones:

Green (stable or improving): AHI change of <1 event/hour/year, SpO2 nadir stable or rising, sleep efficiency stable or above 85%, REM percentage holding at 20-25%.

Yellow (monitor closely): AHI increasing 1-3 events/hour/year, SpO2 nadir trending down toward 90%, sleep efficiency slipping below 85% without an identifiable temporary cause (illness, new medication, major life stress), or REM percentage below 18%.

Red (act now): AHI increase >3 events/hour/year, SpO2 nadir below 88% on any repeat study, or new severe OSA (AHI ≥30) at any life stage, especially during pregnancy.

Repeat polysomnography is not routinely ordered annually in all women, but the AASM recommends retesting when symptoms worsen, when significant weight change occurs (>10% body weight in either direction), or when a woman moves through a major hormonal transition such as menopause.

How to Track Your Results Over Time

Request a copy of your full PSG report, not just the summary letter. The report should include a hypnogram (a visual graph of your sleep stages across the night), the AHI broken down by body position and by sleep stage, and the SpO2 histogram. Compare each metric across studies rather than relying on a single "pass/fail" read.

If your AHI is higher in supine (lying on your back) position than in lateral position, that is positional OSA, a pattern more common in women with milder overall AHI scores. Research in the Journal of Clinical Sleep Medicine found that positional therapy alone controls OSA in roughly 56% of patients with positional-predominant disease.

Normal Ranges and What "Optimal" Actually Means for Women

"Normal" on a sleep study is a population-derived statistical threshold. "Optimal" is a higher bar aimed at the sleep quality associated with the best long-term cognitive, metabolic, and cardiovascular outcomes.

A 2021 analysis in the Journal of the American Heart Association found that women with an AHI of 5-14.9 (technically mild, technically "within normal") had a 42% higher risk of incident hypertension over a 5-year follow-up compared to women with an AHI below 5. That is not a trivial finding. Mild OSA in women is not a benign bystander.

For women specifically, optimal polysomnography targets look like this:

| Metric | Normal (AASM threshold) | Optimal (longevity-medicine target) | |---|---|---| | AHI | <5 events/hr | <2 events/hr | | SpO2 nadir | >88% | >93% | | ODI (4%) | <5 events/hr | <2 events/hr | | Sleep efficiency | >85% | ≥88% | | REM % | 20-25% | 22-25% | | N3 % | 13-23% | ≥18% | | Sleep latency | <30 min | <20 min | | REM latency | 90-120 min | 90-120 min |

The AHI optimal target of <2 events per hour comes from cardiovascular outcomes data. The Sleep Heart Health Study, which enrolled over 6,000 adults including a majority of women over 50, demonstrated that cardiovascular risk rose in a dose-dependent way with AHI, with the steepest slope occurring below an AHI of 15.

REM Sleep and Why It Matters More in Women

REM sleep is where emotional memory processing, fear extinction, and mood regulation happen. Women experience higher rates of anxiety, PTSD, and depression than men, conditions that are bidirectionally linked to REM disruption. Data from the MROS Sleep Study cross-referenced with the SOF Sleep Study (the SOF cohort was all women) found that women with REM sleep below 15% had significantly higher rates of cognitive decline over 5 years.

REM is disproportionately suppressed by alcohol, benzodiazepines, certain SSRIs, and beta-blockers. If you are taking any of these medications, your sleep study should be interpreted with that context noted explicitly in the report.

N3 Sleep and Metabolic Health

Slow-wave sleep is not just restorative. It is metabolically active. During N3, growth hormone secretion peaks, glucose clearance improves, and cortisol is suppressed. A study in Diabetes Care found that selective suppression of slow-wave sleep across three nights reduced insulin sensitivity by 25% in healthy young adults, a magnitude comparable to gaining 20-30 pounds of body fat.

For women with PCOS, this is directly relevant. PCOS is associated with reduced N3 sleep even after controlling for BMI, as shown in a 2020 study in the Journal of Clinical Endocrinology and Metabolism. A sleep study in a woman with PCOS should specifically report N3 percentage and sleep efficiency as part of metabolic assessment, not only OSA screening.

How Hormones Across Your Life Stage Change Sleep Study Results

Sleep architecture is not static. It shifts with your hormonal status in ways that directly alter what your polysomnography shows, what it means, and what should be done about it.

Reproductive Years and the Menstrual Cycle

Progesterone is a respiratory stimulant and also promotes slow-wave sleep. During the luteal phase (days 15-28 of a typical cycle), progesterone peaks and many women actually sleep more deeply and have lower AHI compared to the follicular phase. A study in the Journal of Applied Physiology found that AHI was significantly lower in the luteal phase compared to the follicular phase in women with known OSA.

This matters for scheduling. A sleep study performed in the follicular phase (lower progesterone) may show a higher AHI than one performed in the luteal phase. Standardizing the timing of the test to a specific cycle phase is not current clinical practice, but you should note where you are in your cycle when you receive your results. The variability is real.

Pregnancy

Pregnancy is one of the highest-risk periods for new-onset OSA. The combination of weight gain, upper airway edema from estrogen, progesterone-driven respiratory drive changes, and the mechanical compression of a growing uterus on the diaphragm creates a perfect environment for disordered breathing during sleep.

An ACOG Practice Bulletin from 2021 on obesity in pregnancy notes that OSA during pregnancy is associated with gestational hypertension, preeclampsia, gestational diabetes, and preterm birth. A 2020 meta-analysis in Sleep Medicine Reviews found that OSA during pregnancy more than doubled the odds of preeclampsia.

Polysomnography in pregnancy uses the same AHI scoring thresholds, but the clinical response threshold is lower. Most sleep medicine specialists recommend treating an AHI of ≥5 events per hour in a pregnant woman, compared to the ≥15 threshold sometimes applied in asymptomatic non-pregnant adults.

Home sleep testing (HSAT) has lower sensitivity in pregnancy because it cannot capture sleep position adequately and misses central apneas that are more common in this period. A full in-lab PSG is preferred when pregnancy OSA is suspected.

Postpartum: OSA often improves after delivery as airway edema and weight redistribute, but it does not always resolve. Postpartum sleep fragmentation from infant care also artificially suppresses total sleep time, which can compress AHI by reducing the denominator (hours of sleep recorded). A postpartum sleep study should be interpreted with caution if total sleep time recorded was under 4 hours.

Perimenopause

The perimenopausal transition is the period of greatest OSA risk acceleration in women. As estrogen and progesterone decline, the respiratory-stimulant and airway-tone benefits of these hormones are lost. Data from the Study of Women's Health Across the Nation (SWAN) showed that sleep disturbances increased substantially as women approached menopause, even after controlling for vasomotor symptoms.

A NAMS (The Menopause Society) position statement acknowledges the link between menopause and sleep-disordered breathing and notes that menopausal hormone therapy (MHT) may reduce OSA severity in symptomatic perimenopausal women, though it is not approved as a primary OSA treatment. One randomized controlled trial, Cistulli et al. Published in the American Journal of Respiratory and Critical Care Medicine, found that progesterone supplementation reduced AHI by approximately 30% in postmenopausal women with OSA.

If you are perimenopausal and your sleep study shows an AHI in the mild range (5-14.9) with significant symptoms, declining estrogen and progesterone may be contributing. A repeat study a year later is reasonable, and a conversation about MHT with your provider should consider its potential sleep-related benefits alongside its other risks and benefits.

Post-Menopause

Post-menopause is the life stage at which OSA prevalence in women most closely approaches that of men. The Wisconsin Sleep Cohort found OSA prevalence of approximately 21% in postmenopausal women versus 4% in premenopausal women, a more than fivefold difference. N3 sleep also continues to decline. Many postmenopausal women have N3 percentages below 10%, which falls outside even the broad "normal" range and has direct implications for cognitive reserve and metabolic health.

Who Should Get a Sleep Study: A Life-Stage Guide

Not every woman needs a polysomnography. But the bar should be lower for women than current practice often reflects, because standard screening tools were validated in predominantly male populations.

Consider a sleep study if you are:

• In reproductive years with PCOS, unexplained fatigue despite adequate time in bed, or a BMI >30 with snoring reported by a bed partner
• Pregnant, with any combination of snoring, witnessed apneas, morning headaches, or new-onset hypertension
• Perimenopausal or postmenopausal with waking unrefreshed, new or worsening hypertension, or cognitive symptoms (word-finding difficulty, memory lapses) that began alongside sleep changes
• At any age with treatment-resistant depression or anxiety, as undiagnosed OSA is a common contributor
• At any age with newly diagnosed atrial fibrillation or poorly controlled type 2 diabetes or insulin resistance

A sleep study is less likely to be the next step if:

• Your primary complaint is difficulty falling asleep without other symptoms, which points toward insomnia disorder evaluated by the Insomnia Severity Index rather than polysomnography
• Your fatigue is fully explained by an identified cause being actively treated
• You have no risk factors and your Epworth Sleepiness Scale score is below 8

The STOP-BANG questionnaire (Snoring, Tiredness, Observed apnea, Pressure/hypertension, BMI >35, Age >50, Neck circumference >40cm, Gender male) was validated in surgical populations and assigns male gender as a risk point. A 2016 paper in Anesthesia and Analgesia found that STOP-BANG misclassifies a meaningful proportion of women as low risk. For women, symptoms such as insomnia, morning headaches, nocturia, and mood disturbance should carry more diagnostic weight than STOP-BANG's gender-penalizing scoring.

What a Good Report Should Include: Asking the Right Questions

When you receive your polysomnography report, it should contain more than an AHI number and a recommendation. A report that will genuinely guide your care should specify:

1. AHI by position. Supine versus lateral AHI separately reported.
2. AHI by sleep stage. REM AHI versus NREM AHI. REM-predominant OSA is far more common in women and is missed when only the overall AHI is used.
3. SpO2 histogram. How many minutes were spent below 90%, below 85%, and below 80%.
4. Sleep stage percentages. N1, N2, N3, and REM as percentages of total sleep time.
5. Hypnogram. A visual showing the cycling pattern across the night.
6. Arousal index. Number of EEG arousals per hour. An elevated arousal index with a low AHI suggests upper airway resistance syndrome (UARS), which is more prevalent in women and is often missed when only AHI is screened.
7. Total recording time versus total sleep time. The gap between these two numbers tells you whether the recorded AHI is meaningful or artificially compressed.
8. PLMS (periodic limb movement of sleep) index. If present and >15 movements per hour with arousals, this requires separate evaluation and treatment.

The AASM clinical guidelines for PSG reporting require all these elements. If your report is missing them, you can request the full raw data file from the lab.

Pregnancy, Lactation, and Contraception Considerations

Polysomnography is a diagnostic test, not a drug, so there are no direct teratogenicity or lactation concerns with the test itself. The considerations in this section apply to the treatments that follow from a positive result.

CPAP in Pregnancy

Continuous positive airway pressure (CPAP) is safe in pregnancy and is the first-line treatment for OSA diagnosed during pregnancy. A 2018 systematic review in Sleep Medicine Reviews found CPAP adherence in pregnancy to be lower than in non-pregnant populations, largely due to nasal congestion and positional discomfort. Side-sleeping (left lateral decubitus) position is recommended in the third trimester both for fetal circulation and to reduce positional OSA.

ACOG recommends that pregnant women with confirmed OSA be started on CPAP regardless of AHI severity, given the maternal and fetal risks associated with nocturnal hypoxia.

Pharmacologic Sleep Aids and Pregnancy

Several medications sometimes prescribed to improve sleep architecture carry significant pregnancy risks.

Benzodiazepines and Z-drugs (zolpidem, eszopiclone) are classified as having potential fetal risk by the FDA and should be avoided in pregnancy. The FDA label for zolpidem notes neonatal withdrawal and respiratory depression as concerns with third-trimester use.

Melatonin is widely used but has no established safety data in human pregnancy. Animal data suggest potential dose-dependent effects on fetal circadian programming. A 2022 Cochrane review on pharmacological treatments for insomnia in pregnancy found insufficient evidence to recommend any pharmacologic agent as safe for routine use.

If you are pregnant and receiving a sleep study result indicating OSA or significant sleep architecture disruption, the conversation with your provider should prioritize CPAP, positional therapy, and sleep hygiene before any pharmacologic intervention.

Contraception Note for Certain Sleep Medications

Women of reproductive age taking teratogenic or potentially teratogenic sleep medications should use reliable contraception. This applies specifically to sodium oxybate (Xyrem), which is sometimes used for narcolepsy with cataplexy identified on PSG. Sodium oxybate carries a REMS requirement. The prescribing information states it should not be used in pregnancy, and women of reproductive potential should use effective contraception during treatment.

The OSA-Testosterone Connection in Women

OSA in women reduces androgen levels through a different pathway than in men. In men, OSA suppresses testosterone. In women, a 2021 cross-sectional analysis in the Journal of Clinical Endocrinology and Metabolism found that moderate-to-severe OSA was associated with elevated total testosterone and DHEA-S in premenopausal women, possibly through HPA-axis dysregulation and sympathetic nervous system activation. This is directly relevant to PCOS, where OSA and hyperandrogenism co-occur at high rates.

If your sleep study results are being interpreted in the context of testosterone optimization or PCOS treatment, your provider should factor in OSA status as a modifiable variable. Treating OSA with CPAP has been shown to reduce fasting insulin and HOMA-IR in women with PCOS, as reported in a 2012 trial in the European Journal of Endocrinology, making the sleep study one of the most clinically actionable labs in the PCOS workup.