Overnight inhalation of 2.5% CO2 delivered via a novel low-resistance open-mask system enhances sleep efficiency without adverse physiological effects, and concentrations ≤3.5% appear safe with no cumulative effect observed after multiple consecutive nights.
Key Findings
Results
Sleep efficiency was highest at 2.5% CO2 and lowest at 5.0% CO2, with a significant difference between these concentrations.
Sleep efficiency at 2.5% CO2 was 91.7 ± 5.7%, compared to 78.5 ± 10.1% at 5.0% CO2 (P < 0.001).
Sleep efficiency at 3.5% CO2 was similar to room air (0.0% CO2).
Each night involved inhalation of a different CO2 concentration (0.0%, 2.5%, 3.5%, or 5.0%) in randomized order.
Results
Inhalation of 5.0% CO2 reduced REM sleep and increased arousals during the night.
The 5.0% CO2 concentration was associated with reduced rapid eye movement (REM) sleep compared to lower concentrations.
Arousals were increased at 5.0% CO2 relative to other conditions.
These adverse sleep architecture effects were not observed at concentrations ≤3.5% CO2.
The study used polysomnography with diaphragmatic electromyography (EMG) recorded via esophageal electrodes to assess sleep architecture.
Results
Neural respiratory drive increased in a dose-dependent manner with increasing CO2 concentration.
Diaphragmatic EMG recorded via esophageal electrodes was used to measure neural respiratory drive.
The dose-dependent increase in neural respiratory drive was observed across concentrations of 0.0%, 2.5%, 3.5%, and 5.0% CO2.
Participants first underwent polysomnography with EMG under inhalation of different CO2 concentrations to establish this dose-response relationship before the overnight sessions.
Results
Blood gases, blood pressure, heart rate, and catecholamines remained normal at CO2 concentrations ≤3.5%.
Arterial blood gases were sampled in the evening before inhalation and the following morning during CO2 exposure.
Overnight urinary catecholamines were measured as an indicator of physiological stress.
No adverse changes in blood pressure, heart rate, or catecholamine levels were detected at concentrations ≤3.5%.
These findings indicate that CO2 concentrations up to 3.5% did not cause CO2 retention or physiological stress responses.
Results
Baseline blood gases remained normal after multiple consecutive nights of CO2 inhalation, indicating no cumulative CO2 retention effect.
Participants completed four consecutive overnight polysomnography sessions with different CO2 concentrations.
Morning arterial blood gases were assessed after each night of CO2 exposure.
No cumulative effect on baseline blood gases was observed across the four consecutive nights.
The authors concluded that no cumulative CO2 accumulation occurs with repeated nightly exposures in healthy subjects.
Results
2.5% CO2 inhalation increased N3 (slow-wave) sleep in healthy adults.
N3 sleep, also known as deep or slow-wave sleep, was increased at the 2.5% CO2 concentration.
This effect occurred without CO2 retention or physiological stress markers.
The novel low-resistance open-mask system was used to deliver CO2 during sleep.
This finding is highlighted as a new and noteworthy result supporting the potential therapeutic utility of low-dose CO2.
Background
A novel low-resistance, minimal dead-space mask was developed to deliver CO2 during sleep, addressing tolerability limitations of existing delivery systems.
Existing CO2 delivery systems introduce dead space and resistance that limit tolerability, motivating development of the new device.
The novel mask is described as a 'low-resistance open-mask system.'
The device was designed to deliver controlled CO2 concentrations of 0.0%, 2.5%, 3.5%, and 5.0% during overnight sleep.
The system was evaluated for its effects on neural respiratory drive, sleep architecture, and potential CO2 accumulation.
Conclusions
The authors support the potential of low-dose CO2 delivery for treating hypocapnia-related central sleep apnea based on these physiological safety findings.
Low-dose CO2 is described as capable of stabilizing ventilatory drive, reducing sleep-disordered breathing, and improving sleep quality.
The study population was healthy volunteers (n=16), not patients with sleep-disordered breathing.
The findings are framed as demonstrating 'physiological safety of low-dose CO2 during sleep.'
CO2 up to 3.5% was described as 'safe and well tolerated' with no cumulative effects.
What This Means
This research suggests that breathing a small amount of extra carbon dioxide (CO2) through a specially designed mask during sleep can improve sleep quality in healthy adults. The study tested four different CO2 concentrations (0%, 2.5%, 3.5%, and 5.0%) across four consecutive nights in 16 healthy volunteers. The key finding was that inhaling 2.5% CO2 produced the best sleep efficiency—about 92% compared to roughly 78% at the highest dose of 5.0%—and also increased time spent in deep (slow-wave) sleep. At the highest concentration tested (5.0%), sleep was disrupted, with more awakenings and less REM (dreaming) sleep.
Importantly, the study found that breathing CO2 at concentrations up to 3.5% did not cause CO2 to build up in the body or trigger stress responses. Blood gases, blood pressure, heart rate, and stress hormone levels all remained normal at these lower concentrations. Additionally, after four consecutive nights of CO2 inhalation, there was no evidence of any cumulative buildup of CO2 in the body, suggesting repeated nightly use is safe in healthy people. The novel mask used in the study was specifically designed to minimize breathing resistance and extra airspace (dead space), which had been problems with older CO2 delivery devices.
This research matters because some forms of sleep apnea—particularly central sleep apnea—are linked to abnormally low CO2 levels that destabilize breathing during sleep. This research suggests that delivering small, controlled amounts of CO2 through a well-designed mask could potentially be a safe and effective approach for treating these conditions, though the current study was conducted only in healthy volunteers and further research in patients would be needed to confirm therapeutic benefits.
Liang S, He B, Wellman A, Chen Y, Wang Y, Deng X, et al.. (2026). Effects of controlled carbon dioxide delivery by a novel device on sleep in healthy subjects.. Journal of applied physiology (Bethesda, Md. : 1985). https://doi.org/10.1152/japplphysiol.01053.2025