Why You Wake at 2:17 a.m.

For 30 years, Catharina woke up 5 to 10 times every night. 

It started in her twenties. Not with a dramatic event, not with a diagnosis: just a slow unravelling of something that had always worked. She would fall asleep, then surface. Fall asleep, surface again. All night, every night, for three decades. 

During the day she dragged herself through work hour by hour. When she got home, she collapsed on the sofa. Evenings were not for living. They were for recovering from the effort of being awake. She tried sleeping pills. They helped her fall asleep, but the sleep they gave her never truly restored her. She was caught between dependency and exhaustion, and the years kept passing. 

Then, at 50, she taped her mouth shut before bed. 

When she wrote to me, she said she had slept through more full nights in four weeks than in the previous thirty years. 

Nothing had changed in her bedroom. Nothing had changed in her stress levels or her schedule. What changed was the chemistry inside her body while she slept. Specifically, her tolerance for carbon dioxide, and what that tolerance determines about the stability of her breathing through the night.

Sleep has a chemistry

When you fall asleep, your breathing slows and becomes more rhythmic. As it does, carbon dioxide builds up in the blood. Research spanning several decades shows that blood CO₂ rises by approximately 5 to 15 percent above waking levels during healthy sleep [1,2,3]. Blood vessels in the brain widen. Neural activity quiets. The brain, reading those signals, begins to allow itself to disengage. 

This is not incidental. The rising CO₂ is part of how sleep becomes possible. It tells the brainstem the situation is stable, ventilation can ease, you can stand down. 

During wakefulness, many things shape breathing: posture, speech, emotion, conscious effort. During sleep, those influences are withdrawn entirely. What remains is the brainstem, managing breathing automatically according to deeply set chemical thresholds. The brain becomes more tolerant of elevated CO₂, accepting a level it would not tolerate while awake, in order to sustain the conditions that deep sleep requires. 

This shift requires trust. Rising CO₂ is ordinarily a danger signal: it can mean the airway has closed or breathing has stopped. For sleep to happen, the brain must interpret that same rise as safe rather than threatening. That reinterpretation is the basis of sleep. And it is exactly where things can go wrong.

Two thresholds, and the space between them 

To understand why so many people wake at 2 a.m., it helps to know that CO₂ regulation during sleep involves not one threshold but two, and they work in opposite directions. 

The first is the apneic threshold. If CO₂ drops too low during sleep (from mouth breathing, overbreathing, or a run of shallow ventilation), the brainstem reads the signal as: no need to breathe right now. Respiratory drive falls. Breathing briefly pauses or becomes very shallow. This is not dramatic; the person rarely knows it happened. But CO₂ is no longer being exhaled fast enough to compensate, so levels begin to rise again. 

Then comes the correction. CO₂ accumulates, rises above the normal sleeping level, crosses a second threshold, the arousal threshold, and the brain pulls the sleeper back to the surface. Sometimes all the way to full wakefulness. More often just enough to fragment the architecture of sleep without leaving a clear memory of having woken at all. 

This is the instability loop that underlies the 2 a.m. waking pattern. It is not simply that CO₂ gets too high and wakes you. It is that CO₂ first drops too low, breathing destabilises, and then the correction (CO₂ rising back up) crosses the arousal threshold. The person surfaces, often confused about why, often believing the cause was a dream, a noise, or a full bladder. 

What determines how vulnerable a person is to this loop is the size of what researchers call the CO₂ reserve: the buffer between normal sleeping CO₂ and the apneic threshold [4]. A wide reserve means the system can absorb small perturbations in breathing without crossing into instability. A narrow reserve means very little is needed to start the cycle. 

What narrows the buffer 

The CO₂ reserve is not fixed. It is shaped by daytime habits as much as by anything that happens during sleep. 

Mouth breathing during sleep. Air moves faster through an open mouth. CO₂ is exhaled more quickly, and the gentle rise that should accompany deepening sleep never builds properly. A common belief is: “I wake up with my mouth closed, so I must be nasal breathing through the night.” Sleep laboratory research using infrared cameras tells a different story. People regularly switch to mouth breathing during the night, particularly when muscle tone drops in deeper sleep stages, without any awareness of it. The mouth that is closed when you wake may have been open at 2 a.m. 

Chronic overbreathing during the day. This keeps daytime CO₂ levels chronically low. The brainstem adapts to this as its new baseline. By the time sleep arrives, there is less CO₂ to rise from, less buffer before the system crosses into instability, and the apneic threshold is closer than it should be. 

Stress and nervous system activation. A nervous system primed for threat is more chemosensitive: it responds to smaller changes in CO₂ with larger ventilatory responses. This increases what researchers call loop gain: the tendency of the breathing control system to overcorrect [4]. A high-loop-gain system is an unstable system. Small perturbations get amplified rather than absorbed. 

Sedentary behaviour. Exercise is one of the most powerful natural stimuli for widening the CO₂ reserve, because it exposes the brainstem to sustained high CO₂ in a way it has to adapt to. During vigorous exercise, CO₂ production rises ten to twenty times the resting rate. Without regular physical activity, that adaptation does not happen, and the brainstem remains poorly calibrated for the CO₂ rises that sleep requires.

When doing everything right isn’t enough

Perhaps you have been practising Conscious Breathing for years. You tape your mouth at night. You breathe slowly through your nose during the day. You understand CO₂. And you still wake at 2 a.m. 

This is not uncommon, and it does not mean the approach is wrong. It points to a distinction that is easy to miss: the difference between the voluntary layer of breathing and the automatic layer. 

During waking hours, breathing is partly under conscious and habitual control. With practice, you can learn to breathe more slowly, more diaphragmatically, and more consistently through the nose. The nervous system adapts. Daytime CO₂ levels rise. Tolerance builds. 

But during sleep, voluntary control is withdrawn entirely. The brainstem takes over, operating according to set points that are older, more conservative, and shaped by different inputs than waking breathing. These set points do not automatically inherit the gains made during the day[5]. The daytime CO₂ baseline you carry into sleep raises the floor of your buffer. But the brainstem’s own calibration (its sensitivity at the apneic threshold) changes more slowly and through different mechanisms. 

Mouth taping helps by reducing CO₂ loss during the night. But if the brainstem’s set point is still calibrated to treat low CO₂ as normal, breathing rate will adjust to compensate. The system finds its way back to the set point. 

What shifts the brainstem set point over time is sustained exposure to higher CO₂ levels through consistent nasal breathing, regular moderate exercise through the nose, and a nervous system given genuine conditions to recover. The process takes months, not weeks. But it does move. 

A few things you may not have considered

Waking to urinate. Many people assume the need to urinate caused them to wake. Research suggests it is frequently the other way: a brief arousal from breathing instability comes first, and awareness of the bladder follows only because the person is already awake[6]. Treating the sleep disruption often resolves the nocturia without any bladder intervention. 

Deliberate hyperventilation practices. If you regularly practise a method that involves intense, forceful breathing (repeated cycles of deep hyperventilation followed by breath-holds), it is worth knowing that chronically low CO₂ can condition the brainstem to treat that low level as normal[7]. For someone already dealing with a narrow CO₂ reserve and disrupted sleep, it may be worth asking whether the practice is supporting the goal or working against it. 

Mouth breathing during exercise. Exercise is one of the main tools for widening the CO₂ reserve. But mouth breathing during exercise produces higher ventilation rates and greater CO₂ loss than nasal breathing at the same workload [8]. The stimulus that should be training the brainstem to tolerate rising CO₂ is being exhaled away. Exercise done through the nose, even if this means slowing the pace at first, is substantially more useful for sleep than the same effort done through the mouth. 

The anxiety loop at 2 a.m. Waking and immediately beginning to monitor (calculating remaining hours, checking whether sleep is returning) is itself a nervous system activation. It narrows the CO₂ reserve from the arousal side. This is not a mindset problem but a physiological one: anxiety-driven overbreathing lowers CO₂, which makes the next sleep cycle harder to sustain. Understanding this loop is the first step to breaking it.

What this means in practice 

Three things consistently move the needle, and the order matters. 

Tape your mouth during sleep. This is the most direct way to prevent the CO₂ drops that start the instability loop. It also removes the question of whether you are switching to mouth breathing during the night without knowing it. 

Breathe lighter and slower through your nose during the day. The CO₂ baseline you carry into sleep is the floor of your buffer. A higher baseline means more room before the apneic threshold. Using a breathing retrainer to slow the exhale and reduce the overall breathing rate builds this steadily over time. 

Exercise through your nose. Not for the tiredness it produces, but for the brainstem adaptation it creates. Daily nasal walking or nasal training, at whatever pace allows comfortable nasal breathing, shifts the brainstem set point in a way that daytime breathing practice alone cannot fully replicate. 

The night itself 

I want to add something personal here. Even now, knowing everything I know about this, there are nights when I lie down and sleep does not come straight away. The old response is to start wondering what is wrong, to get restless, to let the mind start working against itself. 

But understanding the mechanism changes that completely. When sleep does not come, I remind myself: the CO₂ simply is not there yet. The chemistry has not crossed the threshold. My body is doing exactly what it is supposed to do. I do not need to fix anything or force anything. I just need to relax and wait for the biology to catch up. 

The anxiety about not sleeping is itself a nervous system activation. It raises the arousal threshold, makes CO₂ harder to build, and pushes the next sleep cycle further away. Knowing that CO₂ simply needs time to build allows you to relax into the process rather than fight it. And that relaxation, in turn, helps it build more quickly. 

The knowledge is not just theoretical. It is something you can use tonight. 

What comes next in this series 

This is the first of three articles on sleep and CO₂. 

The next looks at something discovered only in 2012: a cleaning system inside the brain that flushes out metabolic waste, including proteins associated with Alzheimer’s disease, during deep sleep. CO₂ is the signal that turns it on. 

The third is a first-person account: what ten days of forced mouth breathing did to Anders’s adrenaline levels, mood, cravings, and sleep quality, and what happened the night it stopped.

References

[1] Midgren B, Hansson L. Changes in transcutaneous PCO₂ with sleep in normal subjects and in patients with chronic respiratory diseases. Eur J Respir Dis. 1987;71(5):388–394.

[2] Douglas NJ, White DP, Weil JV, Pickett CK, Zwillich CW. Hypercapnic ventilatory response in sleeping adults. Am Rev Respir Dis. 1982;126(5):758–762.

[3] Berthon-Jones M, Sullivan CE. Ventilation and arousal responses to hypercapnia in normal sleeping humans. J Appl Physiol. 1984;57(1):59–67.

[4] Dempsey JA, Smith CA. Pathophysiology of human ventilatory control. Eur Respir J. 2014;44(2):495–512.

[5] Dempsey JA, Veasey SC, Morgan BJ, O’Donnell CP. Pathophysiology of sleep apnea. Physiol Rev. 2010;90(1):47–112.

[6] Pressman MR, Figueroa WG, Kendrick-Mohamed J, Greenspon LW, Peterson DD. Nocturia: a rarely recognized symptom of sleep apnea and other occult sleep disorders. Arch Intern Med. 1996;156(5):545–550.

[7] Jack S, Rossiter HB, Pearson MG, Ward SA, Warburton CJ, Whipp BJ. Ventilatory responses to inhaled carbon dioxide, hypoxia, and exercise in idiopathic hyperventilation. Am J Respir Crit Care Med. 2004;170(2):118–125.

[8] Dallam GM, Kies B. The effect of nasal breathing versus oral and oronasal breathing during exercise: a review. J Sports Sci Med. 2020;19(2):419–427.

[9] Bae H, Park I, Joo EY. Assessment of effects of carbon dioxide exposure on sleep stability in insomnia. J Sleep Med. 2024;21(1):44–50.

[10] Hayashi K, Fujikawa M, Sawa T. Hyperventilation-induced hypocapnia changes the pattern of electroencephalographic bicoherence growth during sevoflurane anaesthesia. Br J Anaesth. 2008;101(5):666–672.

The information in this article is for educational purposes only and does not constitute medical advice. If you have a diagnosed sleep disorder or are currently using medical equipment to manage your sleep, please consult your healthcare provider before making changes to your treatment.