Mean Oxygen Saturation

In a lab sleep study of 775 adults from the general population, people with lower mean oxygen saturation during sleep had smaller volumes in brain regions involved in memory, emotion, and attention. The count of breathing pauses, which is the number most sleep reports lead with, showed no such link.

That nightly average is mean SpO2. It appears on overnight sleep tests, including wearable at-home studies. There is no agreed cutoff yet for a healthy nightly average. Lower average values and closely related overnight oxygen measures have still been tied to thicker blood, heart events in people with diabetes, cancer risk, and shorter survival, so getting your own baseline gives you a reference point as the research develops.

What the Overnight Average Captures

Hemoglobin is the protein in red blood cells that carries oxygen from the lungs to the rest of the body. Saturation is the percentage of hemoglobin loaded with oxygen at any moment. The mean is the average of those readings across your sleep window.

No organ makes this number. It comes from how well your lungs move air, how well your heart pumps, and how much hemoglobin you have, so a low average can trace back to any of them. During sleep, a common cause is obstructive sleep apnea. In that condition the throat narrows or closes again and again, and oxygen dips with each episode.

Some drop at night is normal. Among 33,080 people who wore a wrist device in a large research study, nighttime readings averaged about 0.8 percentage point lower than daytime ones, and the daily rhythm looked similar across groups. A night that falls well below your daytime level is the pattern worth following up.

Event counts and the average answer different questions. One tells you how often your breathing falters. The other tells you how much low-oxygen exposure your body had across the night. Long, shallow stretches and slow drifts can pull the average down without adding many counted events.

Brain Structure and Memory

In the 775-person brain imaging study, lower mean saturation during sleep went with smaller volumes in the hippocampus, amygdala, thalamus, putamen, caudate, and parts of the frontal and parietal cortex. The hippocampus helps form memories.

The other ways of scoring nighttime oxygen showed no link to brain volume. That includes the lowest reading of the night, the share of time spent below 90%, and the two standard event counts. The finding is observational, so it shows an association and can't prove that low oxygen caused the shrinkage.

What this means for you: if you're noticing memory slips, a normal breathing-pause count shouldn't end the conversation. A low overnight average is a reason to look harder at how you breathe during sleep, and that part is treatable.

Thicker Blood From Low Nighttime Oxygen

When oxygen runs low night after night, the body can respond by making extra red blood cells. The result is called secondary polycythemia, and it makes the blood thicker. Among 605 young adults evaluated for sleep apnea, 74.4% of whom had it, the overnight average independently predicted secondary polycythemia, while the breathing-pause count did not. Men were more prone to it than women.

Hemoglobin and hematocrit tracked with how low the average ran in that study. Hematocrit is the share of blood volume made up of red cells. Other routine results move with it too. RDW is a measure of how varied red cells are in size. RDW runs higher in sleep apnea and follows the overnight average. In 259 people with sleep apnea, albumin, neutrophil and monocyte counts, and an inflammation index built from blood counts were each independently linked to lower average oxygen.

What this means for you: if a blood count shows high hemoglobin or hematocrit with no clear explanation, your nighttime oxygen is one of the first things to check.

Diabetes, Heart Events, and Survival

The strongest outcome data come from people with diabetes. In a Sleep Heart Health Study analysis of 5,780 people, including 453 with diabetes, the diabetes subgroup with higher average sleep oxygen had lower all-cause and cardiovascular mortality than those with lower average sleep oxygen.

A related measure points the same way. In 1,255 people with type 2 diabetes, more time spent below 90% saturation at night predicted later major cardiovascular events, independent of other risk factors. That measure is the share of the night spent low, which is a close relative of the average but not the same number. Outside diabetes, a community study of 4,886 adults found that nighttime oxygen saturation was linked to death from any cause.

The picture is less clear in people who already have sleep apnea and high cardiovascular risk. In 2,687 such people, several newer ways of scoring overnight oxygen added little to predicting future events. Lower average and baseline oxygen did track with heart failure and heart attack, but only in simpler analyses that didn't account for other risk factors. When risk is already high for other reasons, one more oxygen measure has less room to improve the prediction.

What this means for you: if you have type 2 diabetes, your nights deserve attention alongside HbA1c. HbA1c shows your average blood sugar over about three months.

Cancer and Nighttime Low Oxygen

Several large studies have tied low nighttime oxygen to cancer. Most of them measured how long oxygen stayed low, such as time below 90%, so they describe related measures rather than the overnight average itself.

Among 8,748 adults referred for sleep apnea testing in France, more nighttime low oxygen went with higher cancer incidence, especially lung and breast cancer. In 33,711 adults, sleep apnea severity and nighttime low oxygen were each independently linked to new cancers. A pooled analysis of studies reached the same conclusion for both cancer risk and cancer death.

Why a Single Night Can Mislead You

The overnight average is easy to collect and easy to distort. These are the factors most likely to send you in the wrong direction:

• Skin tone: when a light-based sensor is used, readings can run slightly high in people with darker skin, and the error tends to grow when true oxygen is low. Hospital studies found this hid real low oxygen more often in Black patients, and missed lows were linked to higher mortality and delayed COVID-19 treatment.
• Night-to-night variation: in older adults without symptoms, near-normal results barely changed from one night to the next, but the further results strayed from normal, the more they bounced around. One abnormal night needs repeating.
• Altitude: sleeping at elevation lowers saturation even in healthy people. Test at the elevation where you usually sleep, or your result will reflect the trip.
• Device and sensor site: results vary with the device and where the sensor sits. A ring oximeter matched a lab sleep study closely in adults evaluated for sleep apnea, but earlobe and forehead sensors were more reliable than finger sensors in one study of people with scleroderma. Compare readings from the same type of device.

Tracking Your Trend

Your overnight average is more useful as a line than a single point. Variation between nights increases exactly when results start to look abnormal, so one low night could be noise, while two or three nights that agree are a finding.

Get a baseline, ideally covering more than one night. If you start treatment for a sleep breathing problem, retest 3 to 6 months later to see whether the average moved. After that, check at least once a year, and sooner if your weight goes up, a partner notices snoring, or you start waking unrefreshed.

A slow downward drift over several years is worth acting on even if each reading looks unremarkable on its own. It means something in your breathing, heart, or blood is changing.

What to Do With a Low or Falling Average

Read the average alongside the rest of the sleep report, including the breathing-event counts, the lowest reading, and time below 90%. The combination tells you where to look next.

• Low average with many breathing events: this points to obstructive sleep apnea. Book a sleep medicine evaluation, since the main treatments need a prescription.
• Low average with few events: oxygen may be running low steadily from the lungs or heart. Check your oxygen at rest and during exertion. A one-minute sit-to-stand test can pick up exercise-related drops in people with interstitial lung disease. That condition scars or stiffens the lung tissue. If those results are also low, see a pulmonologist.
• Low average plus high hemoglobin or hematocrit: this fits the thick-blood pattern. Repeat the blood count and deal with the sleep breathing problem.
• Unexpected result with darker skin: assume your true value may be somewhat lower than the reading and act on it sooner.

If you have diabetes or memory concerns, act on a smaller drop than you otherwise would. Those are the groups in which low nighttime oxygen has been tied to the outcomes you'd most want to avoid.