What Comes Back First When You Stop Using CPAP and How Quickly It Happens

What happens when you stop using CPAP, night by night

What happens when you stop using CPAP runs on two clocks. Your apnoea comes back almost at once. Your cardiovascular numbers drift over weeks, not days. Randomised withdrawal trials put breathing events back near untreated levels within the first nights, with clinical relapse at a median of 2.9 days (Sleep and Breathing, 2025). Morning and home systolic blood pressure need roughly two weeks to climb 8.5 to 9.0 mmHg (American Journal of Respiratory and Critical Care Medicine, 2011; Chest, 2016).

That gap between tonight and a fortnight is the entire decision. It separates a planned break from a permanent stop, and it is why coming off CPAP after weight loss is settled with a repeat sleep study rather than a hunch. Here is what the evidence shows, and on which night.

Two things most guides about stopping CPAP get wrong

The first mistake is the popular line that apnoea returns in full force after a few days. It is uncited, and it is late. A 2025 randomised trial in Sleep and Breathing followed 22 patients for one week off therapy. Three measures rose significantly inside that week: the apnoea-hypopnoea index (AHI, the number of breathing pauses per hour), the oxygen desaturation index (ODI), and T90, the time spent with oxygen saturation below 90%. Clinical relapse arrived at a median of 2.9 days in the intermittent-CPAP group and 3.5 days with no CPAP at all. Age and neck circumference predicted who relapsed fastest.

Ask what happens when you stop using CPAP and most pages answer with a mood rather than a measurement. The second mistake is bigger. Nearly every guide treats how you felt after one night off as a personal test result. It is noise. Across two separate four-night withdrawal periods in 25 patients (Pulmonary Therapy, 2021), mean ODI did not differ significantly from night to night (p=0.29 and p=0.40). Yet 14 of those 25 patients, or 56%, changed OSA severity category between the two periods, and 3 shifted into the mild band.

Read that again. The same person, measured twice, landed in two different severity categories. A single tolerable night without your machine is not evidence that your apnoea has eased.

What happens when you stop using CPAP for five nights versus two weeks

The damage curve is not a straight line, and that is the most decision-useful fact on this page. Almost the whole evidence base comes from randomised CPAP withdrawal trials in which patients were switched, blind, to sham CPAP, a machine delivering subtherapeutic pressure, for five to fourteen nights. Five nights and fourteen nights produce very different numbers.

Two details deserve more attention than the first page of search results gives them. In the five-night trial, the blood-pressure rise held up in women and not in men once results were split by sex (Journal of Hypertension, 2025). Half the audience is routinely handed a risk timeline that was not observed in them at that duration. And the sleepiness rebound is modest on average. Across 132 patients the Epworth score rose 2.5 points, not ten (Journal of Sleep Research, 2021). That same analysis named who relapses hardest: younger patients, those with little residual sleepiness on CPAP, and those whose AHI rebounds most.

Why the pressure climbs, and what provably stays unchanged

The driver appears to be intermittent hypoxia rather than sleep fragmentation. One randomised trial tested that directly. Over 14 nights of withdrawal, adding supplemental oxygen cut the rise in mean systolic pressure by 6.6 mmHg (95% CI -11.3 to -1.9; p=0.008) and diastolic by 4.6 mmHg. It also reduced median ODI by 23.8 events per hour (SOX trial, American Journal of Respiratory and Critical Care Medicine, 2019). Blocking the oxygen dips blunted the blood pressure. That points at hypoxic burden, not arousals, as the mechanism.

The same 2011 trial that recorded the 8.5 mmHg morning rise measured two other things. Urinary catecholamines rose, a marker of sympathetic activation. Endothelial function fell 3.2%, the flow-mediated dilation measure of how well an artery widens (American Journal of Respiratory and Critical Care Medicine, 2011). Untreated obstructive sleep apnoea is separately associated with resistant hypertension and with atrial fibrillation. Two weeks off CPAP lengthened QTc by 21.4 ms, and the change tracked the AHI rebound (r=0.60), a concrete electrical route from a stop to arrhythmia risk (European Heart Journal, 2012).

Now the part nobody publishes, because it reads as less alarming. Plenty of things did not move. The 2011 trial found no significant change in inflammatory or metabolic markers over two weeks. The five-night trial found arterial stiffness unaffected (Journal of Hypertension, 2025). Across the two-week withdrawal trials that looked for them, cerebral vascular reactivity, myocardial perfusion, retinal microvascular function and most oxidative-stress markers were also unchanged.

Stopping CPAP for a fortnight is a real cardiovascular event with a narrow footprint. It is not organ-wide collapse, and pretending otherwise costs credibility with the people who most need to hear the true part.

In Europe, stopping CPAP is also a driving-licence event

This is the piece no American page can tell you. EU Directive 2014/85/EU writes the thresholds into law, not just into clinical guidance. An AHI of 15 to 29 defines moderate obstructive sleep apnoea syndrome, and an AHI of 30 or more defines severe, both when associated with excessive daytime sleepiness.

Under the same directive, a licence for moderate or severe OSAS is granted only to applicants who show adequate control of their condition and compliance with appropriate treatment, confirmed by authorised medical opinion. Stopping therapy removes the documented compliance your licence was granted on. Drivers under treatment face periodic medical review at intervals of no more than three years for group 1 and one year for group 2, so a professional driver who quits faces an annual, documented check. Sleep clinicians can also test daytime vigilance objectively, with tools such as the Maintenance of Wakefulness Test, instead of relying on how alert you say you feel.

A 2025 European Respiratory Journal survey covered 25 of the 27 EU member states plus eight non-member states, roughly 770 million people. Every EU member had transposed the directive into national law, several with stricter criteria, including mandatory minimum treatment periods. It also documented widespread under-reporting by patients who fear losing their licence. In the UK, NHS guidance is blunt. Where sleep apnoea with excessive sleepiness is confirmed, you must not drive until symptoms are under control, and the DVLA may need to be told.

In France, quitting is visible to the payer within days

French PPC (pression positive continue) reimbursement is tied to measured usage, not to a declaration. Under the Assurance Maladie tariff framework, from 2026 the observance threshold is 4 hours per night, which works out at a minimum of 112 hours of monthly use for full reimbursement. Below 56 hours per month the weekly forfait falls to roughly EUR 3.65, against more than EUR 14 for regular use, with the first 13 weeks of adaptation paid at EUR 15.96.

Automatic daily transmission of machine data, the telemonitoring the French system calls télésuivi, is now the norm, and refusing it reduces eligibility. Prescription renewal comes after three years and depends on machine data confirming the ventilation is effective. Treatment eligibility itself is defined as an IAH above 30, or between 15 and 30 where cardiovascular comorbidity or severe daytime sleepiness, somnolence diurne, is present. From 2026, only clinicians holding an FST sommeil qualification or a validated 40-hour pathway may prescribe PPC, which narrows who you can go back to for a restart.

European abandonment figures also differ from the 30-50% range that US pages repeat. A five-year cohort of 637 patients at Grenoble University Hospital recorded CPAP termination in 17.1% at one year, 26.8% at three years and 28.9% at five (Sleep Medicine, 2025). A separate French real-world cohort of 4,675 patients found 26.1% terminated over a median 1.5 years, with 21.8% non-adherent (Archivos de Bronconeumología, 2025). German readers search this as CPAP-Therapie abbrechen, Spanish readers as dejar el CPAP. The physiology is identical across the EU. The paperwork is not.

A four-step plan if you have already stopped

What happens when you stop using CPAP without a replacement is plain enough. You go back to untreated SAOS. The withdrawal trials above studied moderate-to-severe patients, whose AHI rebounds to roughly 30 to 45 events per hour off therapy. The aim here is not guilt. It is a supervised route back to a treated airway.

1. Write down why you stopped. Mask leak, pressure intolerance, nasal blockage, claustrophobia and noise each have a different fix. Nasal obstruction in particular is treatable, and it is often the real reason therapy failed.
2. Get measured, not remembered. Ask for repeat polysomnography or respiratory polygraphy rather than reasoning from one night. Understanding what AHI numbers actually mean makes that conversation shorter.
3. Check your licence position. If your diagnosis put you in the moderate or severe band, confirm what your national authority requires while you are between treatments.
4. Pick a tier that matches your measured severity, then verify it. Every option below needs a check afterwards, not an assumption.

Matching the option to your measured severity

Two mechanisms often get confused in that bottom row. A nasal EPAP valve uses small one-way valves at the nostrils to create back-pressure on exhalation. An intranasal stent works mechanically and continuously, holding the nasal passage open on the in-breath and the out-breath alike. Back2Sleep is a CE-certified Class I soft silicone intranasal stent of that second type, from a French company based in Paris, with no electricity, noise or tubing. It needs no prescription, the starter kit holds four sizes at around EUR 39 paid out of pocket, and it is intended for snoring and mild-to-moderate obstructive sleep apnoea. It is not a replacement for CPAP in the moderate-to-severe group the withdrawal trials studied.

One honest caveat on the whole ladder. Whatever you switch to, the untreated interval is where the risk sits, and that is the argument against simply stopping while you decide. If a night without your machine is genuinely unavoidable, on a trip where it cannot travel with you, some airway support is a more sensible fallback than none. A single withdrawn night was enough to degrade simulator driving in patients who were otherwise fully compliant.

Frequently Asked Questions