Why Your AHI Went Up: What to Check First

Your AHI jumped. You opened CPAP Clarity or myAir, saw a number higher than your usual, and now you're wondering if something is wrong.

Most of the time, a single bad night is just a single bad night. Bodies, sleep, and air are all variable. What matters is the pattern across a week or more, not any one night in isolation.

Here is the order I would check things in.

A Quick Primer on AHI

AHI is the apnea-hypopnea index: the number of breathing-event slowdowns or stops per hour of sleep. Per the AASM Manual for the Scoring of Sleep and Associated Events (current edition is v3, with the v2 1A hypopnea rule still widely used in clinical practice), events are split into four kinds:

• Obstructive apnea (OA): complete airway closure for at least 10 seconds despite continued respiratory effort. The classic OSA event.
• Hypopnea (AASM 1A recommended rule): ≥30% reduction in nasal-pressure airflow for ≥10 seconds, paired with either a ≥3% oxygen desaturation or an EEG arousal. The alternative 1B rule requires a ≥4% desaturation without the arousal option; some labs and Medicare scoring use 1B.
• Central apnea (CA): breathing pause with no respiratory effort. Brain-driven, not airway-driven.
• Mixed apnea: starts central, ends obstructive. Mostly counted as OA in your dashboard.

Your machine reports total AHI, but the type of events that went up matters for triage. Obstructive events suggest a mask, leak, or pressure issue. Central events suggest something different (medication, altitude, or treatment-emergent central sleep apnea). The dashboard breaks the count down so you do not have to guess.

The clinical severity bands every sleep clinician uses: AHI under 5 is normal, 5 to 14 is mild, 15 to 29 is moderate, 30+ is severe. Treatment goal on CPAP is usually below 5 with most providers, though "feeling rested" matters as much as the number.

Step 1: Is this actually a trend, or one rough night?

Open CPAP Clarity (free, runs entirely in your browser) and go to the history page. Look at the last 14 nights. A single spike against an otherwise flat line is normal and usually does not need action. A pattern where your AHI has been climbing for several nights in a row is what we want to investigate.

If it is one rough night, note what was different about that day (new pillow, cold, drinks, travel, stress) and watch the next two or three nights before doing anything.

If the line is trending up, keep going.

The 7-night and 30-night rolling averages on the history page filter out single-night noise. A 30-night average drifting upward is a real signal; a 7-night average bouncing within 2 events per hour is normal.

Step 2: Check your leak first

A mask leak is the single most common cause of a rising AHI. When your mask leaks, the machine cannot deliver the pressure it thinks it is delivering, and obstructive events slip through.

On CPAP Clarity, look at your leak trend. If your leak is above 24 L/min on recent nights, or if it is trending up week over week, the mask is the first place to look. Full guide to fixing CPAP leaks.

Quick fixes to try tonight:

• Make sure the mask is not over-tightened. Tighter is not always better. Most mask leaks actually get worse past a certain tension.
• Check the cushion for wear. Silicone cushions get shiny and hard with time. If yours has a visible worn ring, replace the cushion (opens in new tab).
• Try a different sleep position. Side sleepers often have leaks on the side their mask is pressed into.
• If you are a mouth breather, a chin strap (opens in new tab) or a full-face mask can stop the leak at the source.

Step 3: Check your pressure versus your prescribed max

If your leak is fine but your AHI still rose, look at your pressure P95 (the pressure your machine hit for 5% of the night). If P95 is near the top of your prescribed range (or bumping against it), your machine wants more pressure than it is allowed to give.

On CPAP Clarity, the dashboard shows your pressure P95 next to your average. A rule of thumb: if P95 is within 0.5 cmH2O of your prescription maximum on three or more nights in a row, it is worth asking your provider to raise the max. More on how CPAP pressure settings work.

Step 4: Check your mask age and cushion schedule

Even without a visible leak spike, an aging mask loses seal quality gradually. Standard CMS / Medicare replacement schedules (HCPCS codes in parentheses):

• Full-face mask cushion (A7031): 1 per month
• Nasal mask cushion (A7032): 2 per month
• Nasal pillows (A7033): 2 per month
• Mask frame (A7030): 1 every 3 months
• Tubing (A7037): 1 every 3 months
• Headgear (A7035) and humidifier water chamber (A7046): 1 every 6 months
• Disposable filter (A7038): 2 per month; reusable filter (A7039) every 6 months

If you have been using the same cushion past these intervals, replace it before you change anything else. This is the cheapest fix and often the one that works. Eligible veterans can typically order these on a comparable schedule through VA prosthetics; see How to Order CPAP Supplies Through the VA for eligibility and ordering details.

Step 5: Did your centrals jump?

Sometimes the number going up is not obstructive events but centrals (brain-driven pauses, not airway-related). CPAP Clarity will flag this as "high central apnea rate" or "mostly central events" on the dashboard.

If centrals are up and obstructives are flat, the cause is different; mask changes will not help. This pattern (treatment-emergent central sleep apnea, or TECSA) is common in the early weeks of CPAP therapy, with reported prevalence of 5 to 20% across published cohort studies (aggregate point estimate around 8.4% per Nigam et al.'s 2016 systematic review). In the Liu et al. 2017 Chest trajectory study, approximately 50% of patients with TECSA at week 1 had resolved by week 13 (transient pattern); the rest had persistent or late-emergent centrals. The linked treatment-emergent central apnea article covers the full clinical picture.

If your central count keeps climbing past three months, your provider may consider adjusting pressure, switching to a bilevel device, or evaluating for an underlying cause (heart failure, opioid use, high altitude). What treatment-emergent central apnea is and when it matters.

Step 6: Lifestyle changes in the last 1 to 2 weeks

AHI is sensitive to several things that have nothing to do with your machine:

• Alcohol. Even a couple of drinks the same night can push your AHI up. Effects fade within 24 hours in most people. The mechanism: alcohol relaxes upper-airway muscles and depresses respiratory drive.
• New medication. Opioids, benzodiazepines, gabapentin, and some sleep aids can suppress breathing drive and push AHI (especially central AHI) up. If a prescription changed in the last 2 weeks, that is the most likely cause.
• Illness. A cold, sinus congestion, or any respiratory infection can bump AHI. Allergies (especially seasonal) can do the same by reducing your ability to breathe through your nose. Give it two weeks past recovery before worrying.
• Weight change. A gain of even a few pounds can change your effective pressure requirement, especially if you were already at the top of your range. Per Peppard et al.'s 2000 Wisconsin Sleep Cohort analysis, a 10% body-weight reduction predicts a 26% AHI decrease, while a 10% gain predicts a 32% AHI increase (the relationship is asymmetric: gain hurts more than the equivalent loss helps).
• Position. Sleeping more on your back than usual typically raises AHI. Some people see 2-3x higher AHI on supine nights versus side-sleeping nights.
• Altitude or travel. Sleeping above ~6,500 feet (2,000 m) can raise central apnea events for the first few nights at elevation; the effect varies considerably across individuals and does not always resolve quickly. AutoSet algorithms generally compensate but the residual events can show up.
• Equipment changes. A humidifier turned off (or left on the wrong setting) can dry out your airway and worsen events. A new tube, swapped mask, or changed climate setting all count as "equipment changes" worth noting on the day.

If any of these describe your week, they are the likely cause. You do not need to change settings; you need to let the variable settle.

What NOT to Do

A rising AHI is alarming, but a few specific reactions tend to make things worse:

• Do not change your prescribed pressure yourself. Most modern CPAPs allow it via clinician menu; resist the temptation. Pressure outside your prescription invalidates the therapy and can shift events from obstructive to central or vice versa, making the data harder for your provider to interpret. Talk to your sleep physician first.
• Do not stop using CPAP because of one bad week. A week of AHI in the 7-10 range is still vastly better for your cardiovascular system than untreated OSA at 30+. Treatment continuity matters more than any single night's number.
• Do not buy "CPAP boosters" or pillow attachments marketed as AHI-reducers. None are clinically validated. Every dollar spent there is a dollar not spent on a fresh cushion or a sleep-doctor visit.
• Do not switch masks impulsively. A new mask requires a 1-2 week adaptation period during which your AHI is genuinely worse before it gets better. Make a mask change a deliberate decision tied to a real seal problem, not a panic response to a bad chart.

When AHI Goes Up but You Feel Better

It happens. Sometimes a night with a slightly higher AHI corresponds to better-quality sleep because you spent more time in REM, where AHI is naturally higher (the REM-related muscle atonia includes upper-airway muscles, so OSA is typically worse in REM). A REM-rebound night after a few nights of poor sleep can show a higher AHI yet feel restorative because you finally got the REM sleep your brain was demanding.

If your dashboard shows the AHI is concentrated in the back half of the night (when REM is denser) and you woke up rested, that is biology working as intended, not a regression.

Step 7: When to call your sleep doctor

Book an appointment if any of these are true after you have worked through the steps above:

• Your AHI has been above 5 events per hour for more than two weeks despite troubleshooting.
• Your central apnea rate (not total) is above 5 events per hour for more than three months.
• Your pressure P95 is at your prescribed maximum on most nights and a cushion change did not bring it down.
• You are waking up tired or with headaches despite the data looking normal.
• Anything has changed clinically (new heart condition, new medication, new sleep problem).

Bring your data. How to generate a PDF for your doctor walks through the one-click report from the dashboard. A physician scanning the cover page can see AHI, compliance, dominant event type, and leak quality in under 30 seconds, which makes the conversation much more productive than walking in with a vague "I feel off."

---

This is informational, not medical advice. If you are genuinely worried about your therapy or feeling worse than usual, contact your sleep physician directly. CPAP Clarity surfaces patterns in the data your machine recorded; it does not diagnose or treat.

Sources

• American Academy of Sleep Medicine. AASM Manual for the Scoring of Sleep and Associated Events, v3 (with the v2 1A hypopnea rule still widely used clinically). AASM scoring clarification (opens in new tab)
• Liu D, Armitstead J, Benjafield A, et al. Trajectories of Emergent Central Sleep Apnea During CPAP Therapy. Chest. 2017;152(4):751-760. PubMed Central PMC6026232 (opens in new tab)
• Nigam G, Pathak C, Riaz M. A systematic review on prevalence and risk factors associated with treatment-emergent central sleep apnea. Annals of Thoracic Medicine. 2016;11(3):202-210.
• Peppard PE, Young T, Palta M, Dempsey J, Skatrud J. Longitudinal study of moderate weight change and sleep-disordered breathing. JAMA. 2000;284(23):3015-3021. Source for the asymmetric weight-AHI finding (10% loss → 26% AHI decrease; 10% gain → 32% AHI increase).
• Centers for Medicare & Medicaid Services. CPAP supply HCPCS replacement schedule. Verified against the OIG Medicare DME schedule report. OIG report PDF (opens in new tab)
• Mountain altitude / sleep-disordered breathing literature: Burgess KR et al. and the Mountain West cohort study at PMC3227706 for altitude effects on central apnea events.

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.