Good AHI But Still Tired? Beyond CPAP Data

Your CPAP Says You Slept Great. You Feel Like You Didn't.

You imported your card. Your AHI is 1.8. Mask seal is clean. You used the machine for 7.4 hours. Every number looks like it should. And you still feel like you got hit by a truck.

You are not crazy. Your CPAP machine measures airflow and pressure. It does not measure sleep. The two are correlated but they are not the same thing. A perfect AHI tells you the machine kept your airway open. It does not tell you whether your brain actually rested, whether your blood oxygen stayed stable through the night, or whether your autonomic nervous system was in fight-or-flight mode during what should have been recovery.

This article walks through the specific signals your CPAP data cannot show you, and the layered sources that fill those gaps. The signals are based on the cross-source rules that CPAP Clarity already runs on your data when you connect more than one device. Most users only have CPAP. Once you add a pulse oximeter or Oura ring, the dashboard surfaces the patterns described below automatically. You do not have to find them yourself.

If you have not yet added a second data source and you want a broader read on the lifestyle and clinical factors that can leave a CPAP user tired with good numbers, our Still tired on CPAP walkthrough covers the non-data side of the same question.

The Four Signal Gaps Your CPAP Cannot See

1. Your blood oxygen during the events the CPAP missed

CPAP machines use airflow sensors to detect apneas and hypopneas. Manufacturers set the detection threshold so the machine catches the events it can reliably score. That threshold is conservative. Short events, partial flow limitations, and central apneas that resolve quickly often fall below the threshold and are not counted in your AHI.

A pulse oximeter sees what your CPAP misses. It measures blood oxygen saturation continuously, including during the events your CPAP did not log. The clinical metric is called the oxygen desaturation index, or ODI. It counts the number of meaningful drops in SpO2 per hour. Per the AASM 2018 scoring manual (opens in new tab), a desaturation event is a drop of 3% or more (some labs use 4%) sustained for at least 10 seconds.

If your AHI is very low (under 3) but your ODI is 5 or higher, you have residual respiratory events the machine is not catching. CPAP Clarity surfaces this combination automatically when you import a Wellue O2 Ring (opens in new tab) CSV alongside your CPAP data. The dashboard renders a card called "Desats without obvious apneas" in the Insights section. It compares your ODI to your AHI for the night so you can see the gap.

This is the single highest-yield signal for anyone with persistent fatigue on a normal AHI.

2. Whether you got enough REM sleep

REM sleep is when most memory consolidation, emotional processing, and dreaming happen. Healthy adult sleep architecture spends roughly 20 to 25 percent of total sleep time in REM. If your REM percentage drops below 15 percent on a given night, you may wake up feeling unrested even if every other metric looks fine.

Your CPAP machine cannot measure sleep stages. It only sees airflow and pressure. An Oura ring or similar wrist-worn wearable uses heart-rate variability, body movement, and skin temperature to estimate sleep stage distribution per night. When you connect Oura to CPAP Clarity, the dashboard automatically checks whether REM was suppressed on nights when your CPAP AHI is under 5 (the clean-CPAP gate the rule uses).

Common causes of REM suppression on a clean-AHI night include:

• Alcohol within 4 hours of bedtime
• Late screen exposure or stimulating activity
• Stress and elevated cortisol
• Certain antidepressants. Selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) are widely reported in sleep pharmacology references to suppress REM sleep
• Some pain medications
• Cannabis use

If you see a pattern of low REM on otherwise clean CPAP nights, the most useful next step is a conversation with your prescribing physician (especially for medication review) rather than self-tweaking machine settings that are already correct.

3. Whether your autonomic nervous system was actually at rest

Heart-rate variability (HRV) measures the variation in time between heartbeats. Higher HRV during sleep generally indicates the parasympathetic nervous system (rest, recovery, repair) is in charge. Lower HRV indicates your sympathetic nervous system (alert, vigilant, stressed) is still active.

A wearable like Oura measures HRV continuously through the night and produces a nightly average. CPAP Clarity compares that night's HRV to your personal 30-night baseline.

Two specific night-shapes the cross-source rules catch:

• Autonomic stress. Elevated therapy load (AHI of 5 or higher OR ODI of 5 or higher) combined with HRV in the bottom quartile of your baseline AND at least some time below 88 percent SpO2. The rule reads this as the body staying in fight-or-flight despite ongoing therapy, often because of residual respiratory disruption that is keeping the nervous system on alert.
• Discordant fatigue. Clean CPAP and clean oxygen with a low wearable sleep score. The rule reads this as a non-respiratory cause of poor sleep (stress, illness, recovery debt from previous nights).

Researchers have studied low overnight HRV as one of several variables in cardiovascular risk profiles, particularly in observational sleep apnea cohorts. The mechanism is not a one-line causal claim and the magnitude depends on the population studied. The practical point for you is that a sustained pattern of low overnight HRV is worth bringing to your provider, alongside whatever lifestyle variables (sleep timing, alcohol, stress, evening exercise intensity) you can adjust on your own.

4. Whether your mask leak is disrupting REM specifically

Mask leak is a CPAP number you already see. Your machine reports an average leak rate and a 95th-percentile leak rate per night. What it does not show you is that leak does not affect every sleep stage equally.

When you sleep on your side and shift onto the mask, the seal breaks. The flow disruption typically does not last long enough to register as an apnea, so AHI stays clean and the leak rate average gets diluted across the whole night. But the brief disruption can still fragment REM specifically, because REM is the stage when muscle tone drops and small disturbances are most arousal-prone.

The pattern looks like this. On nights where 95th-percentile leak rises above 24 L/min AND Oura reports REM percentage below 15 percent, the leak is probably the cause of the REM hit even though your AHI is fine. The dashboard catches this combination explicitly through a rule called "leak-and-rem" that fires when both thresholds match on the same night.

For the fix, try a different mask type or a CPAP pillow with cutouts (opens in new tab) for the position you shift into. Your average leak might not need to drop much to bring REM back into normal range.

How to Connect the Layered Data

If you only import CPAP data, you get the CPAP-side analysis. To unlock the cross-source patterns above, add at least one of:

• Wellue O2 Ring (opens in new tab) for continuous overnight pulse oximetry. Wrist-worn, comfortable, exports a CSV that drops into the same import zone. ODI calculation runs automatically and joins to your CPAP nights by date. See the pulse oximeter guide for setup and what the numbers mean.
• Oura ring (Gen 2 or Gen 3) for sleep stage estimation, HRV, heart rate, body temperature deviation, and (Gen 3+) overnight SpO2. Export the account data ZIP from the Oura app and drop it into the dashboard. The Oura article walks through the setup.

CPAP Clarity joins all three sources by date automatically. Once two or more are connected, the cross-source surface on the dashboard activates and the rules described above start surfacing patterns. There is no toggle to flip and no settings page to configure. A synthesis card on the dashboard tells you when your therapy looked clean but something else did not, in plain English, with the specific numbers that triggered the call.

One thing to know about the surface. The cross-source zone needs a recent paired night to render. If your most recent night with two or more sources is more than two weeks old, the dashboard reverts to an empty state that prompts you to reconnect. The rules described in this article all fire only on nights where the second data source is present.

What This Is Not

This article is about closing the gap between "good AHI" and "good sleep" using objective data. It is not a substitute for clinical evaluation. If you are persistently exhausted, your sleep physician should know. Bring the cross-source data with you. A printed therapy report from CPAP Clarity (the "Save as PDF" path from the dashboard) gives them the same view you have.

Three situations where the data is pointing at something clinical, not lifestyle:

• Cross-source synthesis flags a pattern of "concerning" tone for more than 5 nights in 30. That is a trend, not a one-off, and warrants a clinical conversation.
• Your AHI is clean but ODI is consistently 5 or higher across multiple nights. Residual events under the radar are a common reason CPAP users feel tired despite good machine numbers. Your provider can investigate the cause and decide whether a clinical evaluation is appropriate.
• Your wearable shows REM consistently under 15 percent on otherwise clean nights AND you do not match any of the lifestyle causes listed in section 2. Could be medication-related, could be circadian, could be something else worth investigating clinically.

The point is to have the data ready to point at, not to self-diagnose.

Common Questions

Why am I still tired with a good AHI?

A good AHI under 5 means your CPAP successfully kept your airway open through the night. It does not mean you got high-quality sleep. The most common reasons for persistent fatigue with a normal AHI are residual respiratory events the machine did not catch (caught by oximeter ODI), suppressed REM sleep from lifestyle factors or mask leak (caught by a wearable), low heart-rate variability indicating autonomic stress (also caught by a wearable), and non-apnea sleep disorders like periodic limb movements or fragmented sleep. The fastest investigation is to add a pulse oximeter or wearable so the data layers can show you which signal is off.

What does an AHI under 5 actually tell me?

AHI under 5 tells you your CPAP detected fewer than 5 apnea or hypopnea events per hour of therapy. The clinical convention is that under 5 is the therapeutic target for treated obstructive sleep apnea. It does not tell you about events shorter than the machine's detection threshold, about non-respiratory sleep disruptions like periodic limb movements, about sleep architecture quality, or about whether your body recovered. AHI is a useful headline number but it is incomplete on its own.

Should I be worried if my CPAP says everything is fine?

No, but you should be curious. The CPAP says your respiratory therapy is working. Your subjective fatigue is a separate signal. If you feel fine, no investigation needed. If you feel exhausted, the gap between "machine says fine" and "I feel exhausted" usually points at something the machine cannot measure. Layered data tells you which something.

Will a pulse oximeter catch what my CPAP misses?

Often, yes. A consumer-grade overnight pulse oximeter like the Wellue O2 Ring (opens in new tab) measures blood oxygen continuously. If your CPAP-reported AHI is very low (under 3) but your ODI (oxygen desaturation index) is 5 or higher, you have respiratory disruptions the machine is not counting. These are usually short central apneas, partial flow limitations, or hypopneas that fall below the manufacturer's detection threshold. The oximeter sees them; the CPAP does not. CPAP Clarity surfaces this specific pattern as an Insights card on the dashboard called "Desats without obvious apneas".

How do I know if my REM sleep is being suppressed?

You need a wearable that estimates sleep stages, like Oura, Whoop, or Apple Watch. Standard adult REM percentage is 20 to 25 percent of total sleep time. If your wearable consistently shows REM under 15 percent on nights when your CPAP looks clean, REM is being suppressed by something other than the apnea itself. Common causes are alcohol within 4 hours of bedtime, certain antidepressants, late screen exposure, stress, pain medication, and cannabis use.

Why does my mask leak matter if my AHI is still under 5?

Because leak does not affect every sleep stage equally. A leak that briefly disrupts the flow signal may not be long enough to register as an apnea (so AHI stays clean) but is enough to wake you partially during REM specifically, because REM is the most arousal-prone stage. You can have a clean AHI and a sub-clinical REM hit at the same time. The fix is usually a different mask style or a CPAP pillow with cutouts for the position you shift into.

Do I need a sleep study to figure this out?

If you have access to one, yes, a polysomnogram is the clinical gold standard for sleep architecture and respiratory event detection. If you do not have access (cost, insurance, wait time) or you want to see your own data nightly between studies, layered consumer devices fill most of the gap. A take-home sleep test ordered through your physician is a middle option for residual-event screening.

Can CPAP Clarity tell me which factor is the problem?

Yes, automatically, once you connect more than one data source. The dashboard runs cross-source synthesis on every night with 2 or more sources and surfaces a single sentence naming the pattern that fired (or telling you the night looked clean across all sources). The rule set covers residual respiratory events, REM suppression, autonomic stress, mask-leak-during-REM, and several others. You do not configure anything; the synthesis runs on every import.

Bring Your Data to the Conversation

If after reading this you suspect your therapy needs adjustment, the right next step is your sleep provider. Bring the data. CPAP Clarity downloads a therapy report PDF with the relevant numbers (AHI breakdown, leak, pressure, usage, cross-source patterns if connected) so you and your doctor can look at the same view. Self-tweaking pressure settings without a provider is not the recommended path; understanding what your data actually shows so you can ask better questions in your appointment, is.

Analyze your CPAP data and unlock cross-source insights free →

Some links in this article are Amazon Associates affiliate links. As an Amazon Associate I earn from qualifying purchases at no extra cost to you. See our full disclaimer.