Most healthy adults maintain a blood oxygen saturation (SpO2) between about 94% and 97% for the bulk of the night, with brief dips that can touch the low 90s without signaling a problem. A large study of healthy sleepers found an average lowest saturation of around 90%, while the median reading hovered near 96–97%.1PubMed. Normal oxyhemoglobin saturation during sleep. How low does it go? That said, “normal” shifts meaningfully with age, altitude, body position, sleep stage, and underlying health, so a single cutoff doesn’t work for everyone.
What the Numbers Look Like in Healthy Adults
During sleep, your body doesn’t need to fuel active muscles, so oxygen demand drops. But your breathing slows and becomes shallower, especially in deeper sleep stages, so saturation dips a bit compared with wakefulness. In a study of healthy participants, the average SpO2 for the majority of the night (the 50th percentile of recorded values) was about 96.5%, the 10th percentile sat around 94.7%, and the lowest point averaged roughly 90%.1PubMed. Normal oxyhemoglobin saturation during sleep. How low does it go? Brief dips to the high 80s can happen during a position change or a momentary pause in breathing, and in healthy people they resolve within seconds.
Clinicians generally flag sustained readings below 90% as worth investigating. The threshold you’ll see most often in medical literature is “time spent below 90%,” sometimes shortened to T90 or TST90. When that time stretches beyond a few minutes, it starts to matter for cardiovascular and cognitive health, as covered below.
Why SpO2 Drops During REM Sleep
Not all sleep stages treat oxygen the same way. During REM (rapid eye movement) sleep, breathing becomes irregular, shallower, and slower. The intercostal muscles and other breathing-related muscles lose much of their tone, leaving the diaphragm doing most of the work. Upper-airway resistance also increases. The combined effect is that SpO2 tends to run lower in REM than in non-REM stages.2PubMed Central. The Severity of Sleep Disordered Breathing Induces Different Decrease in the Oxygen Saturation During Rapid Eye Movement and Non-Rapid Eye Movement Sleep For a healthy sleeper, the difference is small, maybe a percentage point or two. But if you already have a borderline airway or a lung condition, REM is when oxygen levels are most likely to dip into concerning territory.
This is also why people with obstructive sleep apnea (OSA) tend to have their worst desaturation events during REM. Muscles that ordinarily stiffen the throat relax further, the airway narrows or collapses, and saturation can plunge well below 80% before an arousal restarts breathing.
How Age Changes the Picture
Nighttime oxygen saturation declines gradually with age. A large meta-analysis of polysomnography data found that for each decade of life, mean oxygen saturation during sleep dropped by about 0.6 percentage points, and the minimum saturation dropped by roughly 1.8 points.3The Lancet Respiratory Medicine. Normal values for adult polysomnography: a systematic review and meta-analysis So a 70-year-old with no lung disease might routinely see overnight lows a few points lower than a 30-year-old, and that alone doesn’t mean something is wrong.
An earlier study comparing younger adults (under 45) with older adults found that the older group had more episodes of apnea and low breathing, along with lower oxygen levels, but there was no meaningful difference between men and women in either age group.4American Review of Respiratory Disease. Breathing and Oxygenation during Sleep are Similar in Normal Men and Normal Women The Lancet meta-analysis did find a small sex-related effect, with groups that included more male participants showing slightly lower mean saturations, though the difference per 10% shift in male proportion was only about 0.1%.3The Lancet Respiratory Medicine. Normal values for adult polysomnography: a systematic review and meta-analysis In practical terms, age matters far more than sex.
Altitude Can Rewrite the Rules
If you live at or travel to higher elevations, the thresholds above need recalibrating. With less oxygen in the air, your baseline SpO2 shifts downward even when you’re awake, and sleep widens the gap further. At altitude, the body’s breathing pattern can become unstable: carbon dioxide drops low enough to temporarily shut off the drive to breathe, and breathing only resumes once oxygen falls further. This periodic breathing is extremely common above about 2,500 meters and fragments sleep throughout the night.5PubMed Central. Effects of high altitude on sleep and respiratory system and theirs adaptations
How dramatic can the drop be? In a simulated-altitude experiment (Operation Everest II), researchers tracked oxygen saturation at progressively higher equivalent altitudes. At a pressure mimicking roughly 4,500 meters, average saturation during sleep was about 79%. At higher simulated altitudes, it fell to the 50s and 60s.6American Review of Respiratory Disease. Operation Everest II: Arterial Oxygen Saturation and Sleep at Extreme Simulated Altitude Few people sleep at those extremes, but the principle scales down: even at a modest 1,600 meters, healthy infants have been recorded with a mean SpO2 of about 92–93%.7PubMed. Oxygen saturation by pulse oximetry in healthy infants at an altitude of 1610 m (5280 ft). What is normal? Adults living at similar altitudes see analogous shifts. If you live in Denver, Mexico City, or Bogotá, an overnight reading of 91–93% may be perfectly normal for you.
What Normal Looks Like in Babies
Parents who strap a pulse oximeter on a sleeping infant sometimes panic at readings that would be unremarkable for a newborn. At sea level, healthy full-term infants typically hold a median nighttime SpO2 between 98% and 99% over the first two years of life, and those values were stable across REM and non-REM sleep.8Archives of Disease in Childhood. Nocturnal oxygen saturation profiles of healthy term infants Brief dips are common, though. In one study, about 11% of healthy infants experienced desaturation episodes to around 86%, usually linked to brief upper-airway obstructions during active sleep.9Acta Paediatrica. Breathing patterns, oxygen and carbon dioxide levels in sleeping healthy infants during the first nine months after birth
Altitude compounds the issue for infants just as it does for adults. At about 1,600 meters, the mean SpO2 in healthy infants during sleep was 92–93%, and the lower reference boundary (two standard deviations below the mean) dipped as low as 86% in quiet sleep.7PubMed. Oxygen saturation by pulse oximetry in healthy infants at an altitude of 1610 m (5280 ft). What is normal? So an infant sleeping at altitude with a reading of 88% may be completely fine, while at sea level that same number warrants a call to the pediatrician.
COPD, Heart Failure, and the Conditions That Push SpO2 Lower
Several chronic conditions can drag nighttime oxygen well below the healthy range. COPD is one of the most common culprits. Patients with COPD often show slow, progressive oxygen drops over the course of the night, especially during REM sleep, primarily from shallow breathing rather than airway collapse.10PubMed Central. Respiratory disorders during sleep in chronic obstructive pulmonary disease When COPD overlaps with obstructive sleep apnea, the burden increases: one study found that COPD alone added roughly 46 minutes to the time spent at or below 88% SpO2 during the night, OSA added about 10 minutes independently, and having both conditions along with periodic limb movements of sleep resulted in the greatest hypoxemic burden of all.11PubMed Central. Nocturnal hypoxemia in COPD: the amplifying effect of comorbid OSA and PLMS on oxygen desaturation
Heart failure introduces a different mechanism. Cheyne-Stokes respiration, an abnormal pattern in which breathing cycles between rapid deep breaths and complete pauses, is common in decompensated heart failure and certain neurological conditions. Those apnea phases cause recurring oxygen desaturations and can trigger heart rhythm disturbances.12PubMed Central. Cheyne-Stokes respiration in congestive heart failure
Why Low Nighttime Oxygen Matters for Your Heart and Brain
It’s one thing to tolerate a brief dip; it’s another to spend hours in a low-oxygen state night after night. The cardiovascular consequences are increasingly well documented. When sleep apnea causes intermittent oxygen drops, the body responds with surges of adrenaline-like activity, spikes in blood pressure, and stress on blood vessel walls. Over time this can produce chronic daytime sympathetic overactivity, even during waking hours, amplifying cardiovascular risk well beyond what happens in the bedroom.13PubMed Central. Sleep, Neural Circulatory Control, and Cardiovascular Disease: A Mechanistic Review
Researchers have tried to quantify that risk. One approach uses “hypoxic burden,” a measure that captures how deep and how long oxygen dips are across the night. A study of OSA patients found that higher hypoxic burden and longer time below 90% were significantly associated with a higher 10-year risk of cardiovascular disease and with evidence of reduced blood flow to the heart muscle.14PubMed Central. Hypoxic Burden and T90% as Predictive Indicators of Cardiovascular Risk and Myocardial Ischemia in Patients with Obstructive Sleep Apnea Other large studies have confirmed that the percentage of the night spent below 90% is a stronger predictor of cardiovascular death than the traditional apnea count alone, though that metric also picks up steady-state low oxygen from conditions like COPD and obesity.15European Heart Journal. The hypoxic burden of sleep apnoea predicts cardiovascular disease-related mortality: the Osteoporotic Fractures in Men Study and the Sleep Heart Health Study
The brain is vulnerable, too. A study of community-dwelling adults with high blood pressure found that higher mean overnight SpO2 was associated with better overall cognitive scores, while more time below 90% was linked to worse performance on tests of memory and executive function, even after controlling for vascular risk factors.16PubMed Central. Nighttime hypoxia affects global cognition, memory, and executive function in community-dwelling individuals with hypertension The relationship was dose-dependent: the more minutes spent in the low-oxygen zone, the worse the cognitive scores.
Alcohol, Flying, and Other Things That Drag SpO2 Down
You don’t need a chronic disease to wake up with poor oxygen numbers. Alcohol is one of the most common everyday culprits. In an early but influential study, a night of drinking nearly doubled the number of oxygen desaturation events and increased apneic episodes fivefold in otherwise healthy men. The effect persisted into the next night, even without any further alcohol.17PubMed. Alcohol increases sleep apnea and oxygen desaturation in asymptomatic men Alcohol relaxes the upper-airway muscles, which is why snoring typically worsens after a few drinks, and the same mechanism deepens oxygen dips.
Combining alcohol with reduced cabin pressure during a long-haul flight makes things considerably worse. In a controlled study, young healthy volunteers who drank alcohol and then slept under conditions simulating airline cabin pressure saw median SpO2 fall to about 85%, and they spent roughly 200 minutes below the 90% threshold. Even without alcohol, the cabin-pressure simulation alone dropped median SpO2 to about 88%, with about 173 minutes below 90%. Under normal sea-level conditions, with or without alcohol, none of the participants spent any measurable time below 90%.18PubMed. Effects of moderate alcohol consumption and hypobaric hypoxia: implications for passengers’ sleep, oxygen saturation and heart rate on long-haul flights So if you routinely nod off on a red-eye after a glass of wine, your oxygen is likely spending a big chunk of that nap in a range that would set off alarms in a hospital.
Pregnancy and Nocturnal Oxygen
Pregnant women face a unique set of respiratory changes that can nudge nighttime oxygen readings downward. As the uterus grows, it pushes the diaphragm upward, reducing the lung’s residual volume and functional residual capacity. In the supine position, cardiac output can also drop. The net result is a wider gap between the oxygen in the lungs and the oxygen in the blood.19PubMed. Normal pregnancy and oxygenation during sleep Most healthy pregnant women still maintain saturations well within the normal range, but the margin is thinner. Anyone already at risk for sleep-disordered breathing, whether from pre-existing weight, nasal congestion of pregnancy, or other factors, may find that oxygen dips become more frequent and deeper in the third trimester.
How Accurate Is Your Smartwatch or Finger Oximeter?
Consumer-grade oximeters have become commonplace, from dedicated fingertip clips to wrist-worn smartwatches. Their accuracy varies more than you might expect, and knowing the limitations helps you avoid unnecessary worry or false reassurance.
Wrist-based smartwatches measure reflected light through the underside of the wrist, a much harder location to read than a fingertip. One performance evaluation of a popular smartwatch found an overall root-mean-square error of about 2.3% compared to a medical-grade oximeter, with the error climbing to nearly 3% in people with severe OSA who have the most rapid oxygen swings.20PubMed. Performance evaluation of a wrist-worn reflectance pulse oximeter during sleep Another study comparing a smartwatch against full polysomnography found a similar error range, with the watch tending to read slightly higher than the lab reference.21PubMed Central. Performance of a commercial smart watch compared to polysomnography reference for overnight continuous oximetry measurement and sleep apnea evaluation In patients with COPD, the gap can widen further. A comparison between a smartwatch and arterial blood gas (the gold standard) showed a mean error of about −1.8%, but the limits of agreement ranged from roughly −7.4% to +4.9%, meaning any individual reading could be off by quite a bit in either direction.22PubMed Central. Are Smart Watches Really Smart? Comparison of Blood Oxygen Saturation Values Measured by Smart Watch, Pulse Oximetry and Arterial Blood Gases in Patients with Chronic Obstructive Pulmonary Diseases
What does that mean in practice? If your watch shows 95%, your actual saturation could realistically be anywhere from 92% to 98%. Trends over many nights are more informative than any single snapshot. A pattern of steadily declining readings, or frequent dips into the mid-80s, is worth raising with a doctor even if individual readings are imprecise.
Skin Pigmentation and Pulse Oximeter Bias
One source of error that rarely makes it into consumer marketing materials is skin pigmentation. A comprehensive review of the literature found that pulse oximeters frequently overestimate SpO2 in Black adults and infants, and more broadly in people with darker skin.23PubMed Central. A review of the effect of skin pigmentation on pulse oximeter accuracy The overestimation means a displayed reading of 94% could mask a true saturation below 90%, a phenomenon called occult hypoxemia. This affects both fingertip and wrist-based devices. The FDA has acknowledged the issue, but corrective hardware changes have been slow to arrive. If you have darker skin and rely on home oximetry for a chronic condition, it’s worth flagging this with your clinician. A reading you consider “fine” may warrant more scrutiny than it would for someone with lighter skin.
Overnight Oximetry as a Diagnostic Shortcut
Full in-lab sleep studies (polysomnography) remain the gold standard for diagnosing sleep-disordered breathing, but they’re expensive, labor-intensive, and often involve a long wait. Overnight pulse oximetry, by contrast, can be done at home with a simple finger sensor, and it captures continuous oxygen data throughout the night.24PubMed Central. The uses of overnight pulse oximetry One metric that has gained traction is the oxygen desaturation index (ODI), the number of times per hour that SpO2 drops by a certain amount, typically 3% or 4%. A study comparing ODI against the standard apnea-hypopnea index found strong agreement, with an ODI above 20 showing high sensitivity for identifying severe obstructive sleep apnea.25PubMed Central. Oxygen desaturation index as alternative parameter in screening patients with severe obstructive sleep apnea
Overnight oximetry won’t tell you everything a polysomnogram can. It misses airway-narrowing events that cause arousals without significant oxygen drops, and it can’t differentiate obstructive from central apneas. But as a first-pass screening tool, especially when access to sleep labs is limited, it carries real clinical value.
Supplemental Oxygen for Sleep Apnea
If nighttime oxygen is the problem, why not just add more oxygen? The idea is intuitive, but the evidence is more complicated. A systematic review of supplemental oxygen therapy in obstructive sleep apnea found that oxygen did improve saturation levels compared with placebo across multiple studies. However, the average duration of apnea and hypopnea episodes actually increased in patients receiving oxygen, meaning the breathing pauses lasted longer even though the blood oxygen didn’t fall as far.26PubMed Central. Obstructive sleep apnea and oxygen therapy: a systematic review of the literature and meta-analysis The concern is that better oxygenation may blunt the body’s urgency to resume breathing, removing the safety trigger that ends each apnea. This is why CPAP and oral appliances, which keep the airway physically open, remain first-line treatments for OSA. Supplemental oxygen is used in specific scenarios, such as persistent hypoxemia despite CPAP, overlap with COPD, or heart failure with Cheyne-Stokes breathing, but it’s not a simple fix for a collapsing airway.