Fewer than five events per hour is the widely accepted target for CPAP therapy, and most sleep physicians consider any reading under that threshold a sign that treatment is working well. But that single number hides a surprising amount of nuance: your machine’s event count may not match what a sleep technologist would score by hand, certain types of residual events matter more than others, and a few nights above five does not necessarily mean your therapy is failing.
The Five-Per-Hour Standard
The benchmark comes from the same scale used to diagnose sleep apnea in the first place. An apnea-hypopnea index, or AHI, below five events per hour is considered normal breathing. When you start CPAP, the goal is to push your AHI back down into that range. Many well-treated users see numbers between zero and two on most nights. If your machine consistently reports an AHI under five, your therapy is generally doing its job.
Getting to that target depends partly on how much of the night you actually wear the device. Research has shown that CPAP needs to be worn for roughly two-thirds to five-sixths of total sleep time to keep the AHI below five, assuming the device eliminates events completely while it is on.1PubMed Central. Reliable calculation of the efficacy of non-surgical and surgical treatment of obstructive sleep apnea revisited If you pull the mask off after four hours of an eight-hour night, the untreated hours drag your average up. This is one of the most common reasons people see a number that looks higher than expected.
Your Machine Probably Overcounts
One thing that surprises many CPAP users is that the event count on your machine’s screen or app is not the same as what a trained scorer would find on a full sleep study. CPAP devices estimate events using airflow signals from their internal sensors, and they tend to err on the side of counting more events rather than fewer. In one study comparing the machine’s built-in scoring to manual review by a technologist, the device-reported AHI averaged about 1.5 times higher than the manually scored AHI. The gap was driven almost entirely by hypopneas: the machine flagged roughly twice as many as a human scorer would.2PubMed Central. Accuracy of Positive Airway Pressure Device—Measured Apneas and Hypopneas: Role in Treatment Followup
A separate study found that when using the machine’s automatic scoring, about a third of patients appeared to have a residual AHI above five. But when the same data was scored manually, that proportion jumped to over 60%, because the manual scorer identified events the machine had missed alongside ones it had flagged.3Journal of Clinical Sleep Medicine. Residual Events during Use of CPAP: Prevalence, Predictors, and Detection Accuracy In that data set, the automatically detected average AHI was about 4.4 per hour while the manually scored average was about 7.3. So your machine can both overcount certain types of events and undercount others, depending on the breathing pattern.
The practical takeaway is that you should not panic over a single night reading of, say, 6.2 events per hour. The machine’s estimate is useful as a trend indicator, not a laboratory-grade measurement. Look at patterns across weeks rather than fixating on any one night.
Not All Machines Score Events the Same Way
If you switch from one CPAP brand to another, your reported AHI may change even if your actual breathing has not. The algorithms these devices use to detect and classify respiratory events are proprietary, vary between manufacturers, and even differ across models from the same company.4Archivos de BronconeumologÃa. Performances of Auto-CPAP Devices Under Real-Life Leak Patterns: A High-Fidelity Modeling Approach Clinicians face the same frustration: the numbers from one brand’s report are not directly comparable to another’s, and the way each device estimates and reports air leak adds yet another variable.5Sleep and Breathing. Auto-adjusting positive airway pressure: the fine line between engineering and medicine
This becomes relevant if you read online forums where users compare their nightly numbers. Someone reporting an AHI of 1.2 on one brand’s machine and someone reporting 3.0 on another might be breathing identically. The difference is in how the software counts. If you are comparing your numbers to someone else’s, make sure you are at least on the same platform before drawing conclusions.
Why Residual Events Happen Even With Good Treatment
A perfectly titrated CPAP does not always drive the AHI to zero, and that is normal. Several factors conspire to leave a few events behind each night.
Body position and sleep stage are the biggest culprits. When you sleep on your back, gravity pulls the tongue and soft tissues toward the airway, and the pressure needed to keep things open is higher than when you sleep on your side. During REM sleep, muscle tone drops further, and the combination of being on your back in REM demands the highest pressure of any sleep configuration. Research has found that no CPAP titration should be considered complete unless the patient has slept on their back during REM, because that is the hardest scenario for the device to handle.6PubMed. The sleep supine position has a major effect on optimal nasal continuous positive airway pressure If your device was set based on a study where you mostly slept on your side and did not spend much time in REM, the prescribed pressure may be slightly too low for your worst moments.
Auto-adjusting machines (APAP) try to handle this by raising and lowering pressure throughout the night in response to what they detect. In head-to-head comparisons, auto-titrating and fixed-pressure CPAP produce similar reductions in AHI and other sleep measures.7Systematic Reviews. Auto-titrating versus fixed continuous positive airway pressure for the treatment of obstructive sleep apnea: a systematic review with meta-analyses But auto machines are not perfect either: they react to events after they happen rather than preventing them, so a handful of events will inevitably be scored before the pressure catches up.
Treatment-Emergent Central Sleep Apnea
Some people develop a new type of breathing event after starting CPAP that was not present before. Unlike obstructive events, where the airway physically collapses, central events happen when the brain briefly stops sending the signal to breathe. This is called treatment-emergent central sleep apnea, and it is more common than many users realize.
In a large study tracking CPAP users over their first 13 weeks of therapy, about 3.5% had central apneas appear in either the first or thirteenth week. Among those who did develop them, the pattern was transient in over half of cases, meaning the central events showed up early and then faded on their own. About a quarter had persistent central events throughout, and roughly a fifth developed them later after they were not present initially.8PubMed Central. Trajectories of Emergent Central Sleep Apnea During CPAP Therapy
If your machine or app breaks your events down into “obstructive” and “central” categories and you see the central number creeping up, this is worth discussing with your sleep physician. The transient form usually resolves without any changes, but persistent treatment-emergent central apnea sometimes requires switching to a different device or therapy mode. In refractory cases where CPAP and more advanced auto-servo ventilation are both ineffective, bilevel pressure support with a backup breathing rate is sometimes considered.9Neurotherapeutics. Sleep-Related Breathing Disorders: When CPAP Is Not Enough
How Mask Leak Affects Your Numbers
A leaky mask does more than make annoying noises. When air escapes around the seal, the device’s ability to detect and respond to breathing events degrades, sometimes dramatically. Bench-testing of three popular auto-CPAP machines found that when unintentional leak was introduced, at least one device essentially lost its ability to respond to obstructive apneas and hypopneas correctly. The other two handled certain event types better but still showed reduced accuracy during specific leak scenarios.10PubMed Central. Impact of Unintentional Air Leaks on Automatic Positive Airway Pressure Device Performance in Simulated Sleep Apnea Events
Leak affects your event count in two directions. The machine might undercount events because it cannot detect them through the noise of escaping air. Or it might overcount because the turbulent airflow from the leak mimics the signature of a breathing event. Either way, on nights where your leak was high, your reported AHI is less reliable. Most machines and companion apps flag high-leak nights, and those readings should be interpreted with extra skepticism. If you are consistently seeing high leak numbers alongside higher event counts, the fix is usually a mask adjustment, a different mask style, or a chin strap if you are using a nasal mask and your mouth is falling open.
Does Alcohol Make Your Numbers Worse?
This is one of the more common questions CPAP users have, especially after noticing a spike in events following a night out. Alcohol relaxes the muscles of the upper airway, which is why it worsens sleep apnea in people who are not on treatment. But for those already using CPAP, the picture is less dramatic than you might expect.
A systematic review looking at alcohol’s effects on breathing during sleep found that the limited data on CPAP users suggested alcohol did not significantly change the effective pressure needed to control events. The researchers noted that in patients whose airway is already being held open by positive pressure, alcohol’s additional muscle relaxation did not seem to push things past what the machine could handle.11PubMed Central. The Impact of Alcohol on Breathing Parameters during Sleep: A Systematic Review and Meta-analysis A separate study specifically testing moderate vodka consumption in patients with mild-to-severe sleep apnea found no meaningful change in AHI, oxygen levels, or required CPAP pressure on drinking nights compared to control nights.12European Respiratory Journal. Influence of moderate alcohol consumption on obstructive sleep apnoea with and without AutoSet nasal CPAP therapy
That said, these studies involved moderate amounts of alcohol and relatively small numbers of participants. Heavy drinking is a different story, and alcohol also disrupts sleep architecture in ways that your AHI number will not capture. Your event count might look fine after several drinks, but your sleep quality can still be worse. And practically speaking, alcohol makes it more likely that you will remove your mask in the middle of the night without realizing it, which brings us back to the adherence issue discussed earlier.
When Higher Numbers Actually Signal a Problem
If your AHI is persistently above five despite wearing the mask all night with low leak, that is a genuine signal that something needs attention. The reasons vary: your pressure settings may be too low, your weight may have changed, you may have developed treatment-emergent central events, or your mask may have degraded over months of use. A conversation with your sleep provider is worthwhile at that point, and a follow-up in-lab sleep study with the device on (called a PAP-nap or CPAP titration study) can identify what is happening more precisely than the machine data alone.
The health stakes of untreated residual events are real. Large epidemiological studies show a clear, dose-response relationship between AHI severity and both daytime sleepiness and high blood pressure. Even mild elevations in AHI are associated with a higher percentage of people reporting excessive daytime sleepiness, and the link between AHI and hypertension is substantial: in one cohort, compared to an AHI near zero, an AHI of 15 was associated with roughly 75% higher odds of hypertension, and an AHI of 30 tripled the odds.13Sleep. Metrics of sleep apnea severity: beyond the apnea-hypopnea index
However, there is growing recognition that AHI alone does not capture the full picture. A recent study of patients with coronary artery disease and sleep apnea found that the oxygen-drop burden from events was a stronger predictor of major cardiovascular complications than the AHI itself. Patients with a high oxygen-drop burden had nearly double the risk of cardiovascular events over about five years, while a high AHI on its own was not significantly associated with those events.14CHEST. Association of Hypoxic Burden With Cardiovascular Events In other words, ten brief events with minimal oxygen dips may matter less than three prolonged events that tank your oxygen saturation. Your machine’s AHI counts every event equally, but your body does not experience them equally.
Adherence Versus Event Control
Some CPAP users focus intensely on getting their AHI as close to zero as possible while paying less attention to how many hours per night they actually wear the device. Both matter, but for most people, wearing it consistently matters more than chasing a perfect event score. Analysis of a large French sleep database found that patients whose CPAP was only partially controlling their apnea also tended to use the device for fewer hours per night, averaging about 5.75 hours compared to 6 hours in the group with good event control.15Respirology. Partial failure of CPAP treatment for sleep apnoea: Analysis of the French national sleep database The difference in usage time was modest, but it illustrates a common pattern: the users who are struggling with residual events are often the same ones struggling with adherence, and improving one can improve the other.
If you are getting six or seven hours of wear time with an AHI of two, you are almost certainly getting more benefit than someone who wears the device for three hours but has an AHI of zero during those hours. The untreated portion of the night contributes heavily to overall sleep apnea burden, fatigue, and cardiovascular risk. Prioritize wearing the mask through the full night before worrying about whether your AHI is 1.5 or 2.5.
Night-to-Night Variation Is Normal
Even with steady pressure settings and the same mask, you will see your nightly AHI bounce around. Sleep apnea itself has significant night-to-night variability driven by factors like how much you ate before bed, nasal congestion, sleep position, how much REM sleep you got, and how deeply you slept. On a night where you spend more time in deep REM sleep while lying on your back, your events might tick up. The following night, if you sleep mostly on your side with lighter sleep, the count might drop. A night with a stuffy nose from allergies could push things higher, while a clear-breathing night brings them back down.
This is why sleep physicians look at the trend over 30 days or more rather than a single night’s readout. Most CPAP apps and reports offer a rolling average. If that average sits comfortably below five and you feel rested during the day, the occasional night at six or seven is not a crisis. If the rolling average is creeping upward over weeks, that is a different conversation and might warrant a check of your equipment, settings, or overall health.
What Your Machine Cannot Tell You
Your CPAP’s event report covers apneas, hypopneas, and sometimes flow limitations and central events. What it does not capture includes periodic limb movements, arousals that are not tied to a breathing event, sleep-stage disruptions, and upper airway resistance episodes that fall below the threshold for a scored hypopnea. You can have an AHI of zero on your machine and still wake up feeling unrested because something else is fragmenting your sleep.
If your numbers look good but your symptoms persist, telling your sleep doctor that “my AHI is fine” is useful information but not the whole answer. A follow-up polysomnography, which monitors brain waves, oxygen, leg movements, and many other signals simultaneously, can uncover problems that a CPAP flow sensor will never see. The machine is excellent at its specific job of tracking breathing events, but it is not a comprehensive sleep monitor.