Minimum Oxygen Saturation

If you snore, wake up tired, or have been told you stop breathing in your sleep, one of the most useful numbers from an overnight test is the single lowest saturation reading. In a very large sleep-clinic population followed for five years, people whose overnight low fell further went on to develop an irregular heart rhythm called atrial fibrillation more often, even after accounting for lung function.

The usual sleep apnea score counts how often your breathing stalls each hour. Two people with the same count can have very different nights, since one barely dips and the other drops hard with every pause. Several studies find that the depth and duration of those drops track blood pressure, diabetes, stroke, and survival in ways the count alone misses.

What the Lowest Reading Captures

Hemoglobin is the protein in your red blood cells that carries oxygen. Oxygen saturation is the percentage of it that's loaded with oxygen at any given moment, and an overnight sleep recording can estimate it all night long. The minimum is the single lowest value recorded.

That low point depends on several things working together: how well your lungs pick up oxygen, how well your heart and blood vessels deliver it, and how badly your breathing stalls during sleep. A deeper low generally means a longer or more complete pause in breathing, or lungs that had less reserve to begin with.

Sleep reports usually show the minimum alongside two related numbers. Time below 90% is how much of the night you spent under that line. The oxygen desaturation index counts how many separate drops happen per hour. The minimum describes the worst moment and the other two describe how long and how often, and much of the outcome research draws on all three.

Atrial Fibrillation

Atrial fibrillation is an irregular, often fast heart rhythm that raises the risk of stroke. In the large sleep-clinic cohort mentioned above, someone whose overnight low was 10 percentage points deeper than another person's had about 9% higher risk of developing it over five years. That link held after adjustment for many other factors, including lung function, which suggests the nighttime dips themselves are part of the story.

A French cohort of about 7,200 people referred for sleep apnea testing found something similar using broader overnight oxygen measures. Low nighttime oxygen and an unusually variable pulse rate on the same recording each predicted new atrial fibrillation on their own.

If your low comes back deep, check whether your sleep report includes a suspected atrial fibrillation readout. A deep low plus palpitations, a rhythm flag, or a family history of irregular heartbeat is a good reason to get a longer heart rhythm recording and see a cardiologist.

Stroke, Heart Attack, and Cardiovascular Death

In about 2,900 older men living in the community, those with severe nighttime low oxygen had a higher risk of stroke, and overnight oxygen measures predicted stroke better than the number of breathing pauses did. In a similar group of about 2,800 older men, time spent below 90% predicted death from cardiovascular disease independently of other risk factors.

The pattern also shows up in people who already have heart disease. In a stable heart-failure cohort, nighttime low oxygen independently predicted death from any cause. In people who had a heart attack, moderate to severe sleep apnea and nighttime low oxygen made it more likely that the attack struck between midnight and 6 a.m.

Most of these studies measured how long oxygen stayed low, or the total burden of low oxygen across the night, which is related to the single lowest value but is a different measurement. A deep minimum and a heavy overall burden often appear together, but one brief dip can produce a low minimum with little burden, so it helps to read the numbers side by side.

Blood Pressure, Diabetes, and Kidneys

Among about 365 people with moderate to severe sleep apnea, those whose lowest reading fell very deep and who also spent a large share of the night below 90% had higher odds of high blood pressure and type 2 diabetes, and higher five-year mortality. The researchers described them as a distinct high-risk group that the standard pause count alone didn't pick out.

Related measures point the same way. In a sleep apnea clinic cohort, a lower typical overnight oxygen level predicted type 2 diabetes regardless of pause count, weight, or age. In people who already had type 2 diabetes, a heavier nighttime low-oxygen burden was linked to more major heart events in about 1,250 people and to faster worsening of kidney disease in about 860.

If your blood pressure is hard to control or your blood sugar is drifting upward, treat a deep overnight low as a possible contributor and investigate it. Pairing it with a three-month blood sugar average (HbA1c) and a 24-hour blood pressure recording shows you whether the pieces fit together.

Lung Disease and Scarring

Idiopathic pulmonary fibrosis is a progressive scarring of the lungs with no known cause. In a small study of 31 people with it, a lower overnight low went with shorter survival, more breathlessness, weaker lung function, and higher pressure in the lungs' blood vessels. Their oxygen fell further during sleep than during maximal exercise, so a daytime walking test can miss the worst of it.

In about 50 people with a range of scarring lung diseases, deeper lows on a six-minute walk and overnight, together with more time below 90%, tracked more severe scarring. Nighttime low oxygen has also been tied to worse lung blood-vessel pressures in people with chronic blood clots in the lungs, and people with cystic fibrosis had about three times the odds of moderate to severe sleep-disordered breathing, with lower nighttime oxygen.

Recovery After a Stroke

Among about 90 people in rehabilitation after an acute stroke, a higher overnight low predicted bigger gains in day-to-day independence over their stay. The number of drops per hour did not, which is one example where depth carried information that frequency didn't.

If you or a family member is recovering from a stroke, an overnight oxygen recording during rehab can be useful, because nighttime low oxygen is something that can be found and addressed.

Cancer Risk Is Less Settled

Several large studies of people referred for sleep testing link nighttime low oxygen to cancer. A French cohort of about 8,700 found higher cancer incidence with more nighttime low oxygen, especially lung and breast cancer, and a second clinical cohort of about 34,000 found sleep apnea severity and low oxygen each independently tied to new cancers. A 2021 pooled analysis also found both associated with higher cancer risk and cancer death.

Combined community cohorts tell a different story: once age, sex, and body weight were accounted for, sleep apnea did not raise overall cancer risk or cancer death. The two findings fit together once you consider who was studied. People sent to sleep clinics tend to be older, heavier, and sicker, and those factors raise cancer risk on their own, so clinic studies can't fully separate low oxygen from everything that comes with it. For now, cancer is a possible association with nighttime low oxygen that hasn't been shown to be causal.

How Much Weight to Give the Number

The evidence isn't uniform. In a secondary analysis of a large trial of about 2,700 people with sleep apnea who already carried high cardiovascular risk, newer ways of scoring overnight oxygen added little to predicting heart events, although the pattern of drops was linked to heart failure and heart attack. In a group already at high risk, many factors compete to predict events, so any single overnight number probably adds less.

Standardized cutoffs for the minimum on its own haven't been agreed, devices score it differently, and most outcome data come from people referred for sleep testing. That makes it most useful as one part of the whole overnight picture, and as a number you follow over time.

When Results Can Be Misleading

The minimum comes from a single moment out of thousands, which makes it easy to distort.

• Day-to-day swings: in people with chronic lung disease wearing sensors for a week at home, time spent below 90% varied enough from day to day that whether someone counted as a frequent desaturator changed. One night, and especially one lowest point, can be unrepresentative.
• Movement and poor signal: a sensor that shifts or loses contact can record a drop that never happened, and readings are less reliable when blood flow to the measurement site is poor. Devices that also track motion can help separate dips caused by moving around from breathing-related drops.
• Skin tone and body weight: pulse oximetry studies in hospital and surgical settings show that sensors can read higher than true arterial saturation. The overestimate has been larger in Black patients in several datasets and can also be larger in people with obesity, especially when levels are already low.
• Supplemental oxygen: oxygen can raise the low reading even if the airway is still collapsing or carbon dioxide is rising. A better minimum after oxygen does not prove sleep apnea or underbreathing has been treated.
• Recent illness, altitude, and device type: a recent chest infection or sleeping at altitude can lower readings temporarily, and consumer smartwatches differ widely in accuracy, so compare results from the same device under similar conditions.

Tracking Your Trend

One night gives you a snapshot, and the lowest point in that snapshot is the least stable number on the page. Repeating the recording tells you whether a deep low is your usual pattern or a fluke, and whether it moves after you change something.

A practical cadence: get a baseline overnight recording, repeat it within 3 to 6 months of any meaningful change such as weight loss, starting sleep apnea treatment, or a new mouthpiece, and then repeat at least once a year if you're actively monitoring a known problem. If the first result surprises you, repeat it on another typical night before drawing conclusions.

When you compare, use the same device and read the minimum together with time below 90% and the number of drops per hour. A minimum that stays deep while the other two improve often points to a few severe events still happening, and a steady slide downward over years is worth investigating even if no single reading looks dramatic.

What to Do With a Deep Low

A deep minimum starts a workup. The combination of findings tells you where to look next.

• Deep low with many breathing pauses: this is the classic sleep apnea picture, and paired with long stretches below 90% it marks the higher-risk group described above. Get a full diagnostic sleep study if you haven't had one. A sleep specialist can help choose PAP, an oral appliance, or another treatment.
• Deep low with few pauses, or long flat stretches below 90%: this pattern points more toward the lungs or shallow breathing during sleep. Lung function testing and a pulmonologist come next, and overnight carbon dioxide monitoring can uncover underbreathing at night in people whose daytime results look normal.
• Deep low with a rhythm flag, palpitations, or heart failure: get a longer heart rhythm recording and involve a cardiologist.
• Deep low with high blood pressure or rising blood sugar: check a 24-hour blood pressure recording and blood sugar markers, and treat the sleep finding as part of the same problem.
• One isolated dip on an otherwise clean night: repeat the recording before acting, since a single artifact can set the minimum.

If a sensor shows low oxygen while you're awake and you feel short of breath, confused, or have chest pain, get same-day medical care.