Respiratory Fitness and the Sleep Quality Continuum: A Framework for Training Sleep Through Respiratory Control

Locke Working Paper, May 2026

Abstract

The prevailing clinical framework categorizes sleep into a binary: pathological (diagnosable sleep-disordered breathing) or normal. This paper proposes that sleep quality exists on a continuous, trainable spectrum in which respiratory control variables are an underrecognized and trainable determinant, specifically loop gain, chemoreceptor sensitivity, CO₂ reserve width, cerebrovascular reactivity, and upper airway mechanical stability. These variables, which produce clinically diagnosable sleep apnea at one end of the distribution, also constrain sleep efficiency in ostensibly healthy individuals who would never meet diagnostic criteria for any sleep disorder. The arbitrary nature of the Apnea-Hypopnea Index (AHI) threshold obscures this continuum, creating a threshold artifact that leaves millions of sub-optimal sleepers without explanation or intervention. The paper introduces the concept of 'respiratory fitness' as a trainable physiological trait analogous to cardiovascular fitness: a measurable characteristic that influences sleep quality the way cardiovascular conditioning influences exercise performance. It reviews evidence from altitude physiology, freediving, breath-hold training, inspiratory muscle training, nasal breathing interventions, and pediatric sleep medicine demonstrating that respiratory control variables are modifiable across the entire population spectrum, and that targeted respiratory interventions improve sleep architecture even in non-clinical populations. The practical implication is a paradigm shift from treating sleep pathology to training sleep performance.

The Binary Fallacy

Sleep medicine operates on a diagnostic threshold. An individual with an Apnea-Hypopnea Index below 5 events per hour is classified as normal. An individual at 5 or above is classified as having obstructive sleep apnea. This threshold determines who receives treatment, who receives a diagnosis, and who is told their sleep is fine.

The threshold is arbitrary. Minor changes to hypopnea scoring criteria (requiring 4% versus 3% oxygen desaturation, or including associated cortical arousals) can reclassify the same patient from healthy to pathological. In female cohorts, shifting from 4% to 3% desaturation criteria increases OSA diagnosis rates by 12.8% to 63%. The boundary between 'normal' and 'disordered' is not a physiological discontinuity. It is a scoring convention.

Meanwhile, the respiratory variables that produce sleep apnea at one end of the distribution do not suddenly vanish below the diagnostic threshold. They exist on a continuous spectrum across the entire population. An individual with an AHI of 3 may experience more sleep fragmentation, more daytime fatigue, and worse neurocognitive performance than an individual with an AHI of 10, depending on the relative contributions of flow limitation, respiratory effort-related arousals, and sub-threshold ventilatory instability that the AHI metric was never designed to capture.

This paper presents its arguments at three levels of certainty that the reader should distinguish. First, established physiology: the respiratory variables described exist on continuous distributions, are measurable, and demonstrably affect sleep architecture in clinical populations. Second, plausible extrapolation: these same variables constrain sleep quality in sub-clinical and ostensibly healthy populations, based on convergent evidence from multiple domains. Third, proposed framework: the concept of 'respiratory fitness' as an integrative, trainable construct that predicts and modifies sleep quality across the full population spectrum. The first level is well-established. The second is supported but requires further controlled study. The third is the novel contribution of this paper.

The Respiratory Variables That Govern Sleep Quality

Five measurable respiratory variables determine where an individual falls on the sleep quality continuum. All five exist on continuous distributions in the general population. All five are modifiable.

Loop Gain

Loop gain is the ratio of corrective ventilatory response to the magnitude of the disturbance that triggered it. A system with a loop gain below 1.0 dampens perturbations and returns to stable breathing. A system with a loop gain at or above 1.0 overcompensates, producing oscillatory breathing that fragments sleep. Loop gain is the mathematical product of three components: controller gain (chemoreceptor sensitivity to blood gas changes), plant gain (the efficiency of the lungs in clearing CO₂ per unit of ventilation change), and mixing gain (the circulatory delay between gas exchange and chemoreceptor detection).

Loop gain is not a binary pathological marker. It has been measured on a continuous distribution across clinical and non-clinical populations. In one controlled study, healthy controls showed a loop gain of 0.16 +/- 0.04 at 1 cycle/minute, compared to 0.28 +/- 0.04 in OSA patients. Individuals in large cohorts with sub-clinical AHI below 10 still exhibit significant loop gain values of 0.40 to 0.48. The variable exists on a spectrum, and individuals at the higher end of 'normal' are more vulnerable to sub-threshold ventilatory instability that degrades sleep without producing scoreable events.

CO₂ Reserve and Chemoreceptor Sensitivity

The CO₂ reserve is the difference between the resting eupneic PaCO₂ and the apneic threshold, the PaCO₂ below which the brainstem ceases respiratory drive. During wakefulness, cortical drive overrides the chemical threshold and breathing continues regardless. During NREM sleep, the wakefulness drive is withdrawn and respiration becomes entirely dependent on chemoreceptor control. A narrow CO₂ reserve means any transient hyperventilation, from a sigh, a position shift, or a spontaneous arousal, can drop PaCO₂ below the apneic threshold and trigger a central apnea or a period of ventilatory instability.

The width of the CO₂ reserve is not fixed. It is determined by the baseline PaCO₂ level (which reflects habitual breathing patterns) and by chemoreceptor sensitivity (which determines the slope of the ventilatory response). Both are modifiable through training. Individuals with chronically elevated minute ventilation from stress, anxiety, or habitual mouth breathing operate with a narrowed CO₂ reserve that predisposes them to nocturnal ventilatory instability even if they never meet criteria for any sleep disorder.

Cerebrovascular Reactivity

Cerebrovascular reactivity (CVR) is the brain's hemodynamic buffer against respiratory perturbations. When PaCO₂ drops transiently, robust CVR produces compensatory vasoconstriction that retains CO₂ locally at the brainstem, shielding the chemoreceptors from the systemic fluctuation. When CVR is blunted, the chemoreceptors see every systemic CO₂ change unfiltered, amplifying the destabilizing effect of each ventilatory perturbation.

In experimental models, pharmacologically reducing CVR by approximately 25% (using indomethacin) narrowed the CO₂ reserve from 6.3 +/- 0.5 mmHg to 4.4 +/- 0.7 mmHg and induced spontaneous central apneas in otherwise healthy subjects. In the Wisconsin Sleep Cohort, blunted CVR correlated with mean nocturnal oxygen saturation across the entire spectrum of sleep-disordered breathing, from sub-clinical to severe. CVR is therefore a continuous variable that modulates sleep stability across the population, not a pathological marker confined to clinical populations.

Upper Airway Mechanical Stability

Upper airway collapsibility is the most visible and commonly treated respiratory variable in sleep medicine, but it is only one of five. Airway stability depends on the balance between negative inspiratory pressure (which tends to collapse the pharynx) and dilator muscle tone (which holds it open). When the mouth opens during sleep, the mandible drops, the tongue base retracts, and the pharyngeal airspace narrows. Nasal breathing provides a natural stabilizing effect through physiological positive end-expiratory pressure (PEEP), which pneumatically splints the airway during expiration. Nasal breathing also entrains nitric oxide from the paranasal sinuses, improving ventilation-perfusion matching in the lungs. Inspiratory muscle strength, measurable as maximum inspiratory pressure (MIP), determines the mechanical resilience of the airway dilators against collapse.

Autonomic Balance

The autonomic nervous system mediates the transition between wakefulness and sleep. Sympathetic hyperarousal, whether from chronic stress, anxiety, or the physiological consequences of sub-threshold respiratory events, prevents the consolidation of slow-wave and REM sleep. Heart rate variability (HRV), specifically the high-frequency (HF) power band reflecting parasympathetic vagal tone, is one of the strongest predictors of sleep quality and cardiovascular health. Autonomic balance is both a consequence of respiratory stability (stable breathing promotes parasympathetic dominance) and a modifiable input to it (deliberate slow breathing techniques directly increase vagal tone).

Evidence That the AHI Threshold Is a Diagnostic Discontinuity

Upper Airway Resistance Syndrome

Upper Airway Resistance Syndrome (UARS), first identified by Guilleminault in 1982, embodies the continuum. UARS patients present with AHI well below 5 but suffer from elevated respiratory effort-related arousals (RERAs) producing a high Respiratory Disturbance Index. When directly compared to mild OSA patients (AHI 5-14) and healthy controls, UARS patients report significantly higher chronic daytime fatigue, anxiety, and depression than those with formally diagnosed mild OSA. On psychomotor vigilance testing, UARS patients show more sustained attention lapses than mild OSA patients. A patient with an AHI of 3 can be functionally more impaired than a patient with an AHI of 10.

Sub-Threshold Flow Limitation in 'Normal' Sleepers

Inspiratory flow limitation, where increasing respiratory effort fails to produce proportional airflow due to partial pharyngeal collapse, is prevalent even in individuals with a normal AHI. Studies measuring flow limitation via nasal cannula pressure transducers show a progressive increase in flow-limited breaths across the continuum: approximately 5% of breaths in asymptomatic normal sleepers, 6.1% in asymptomatic individuals with AHI 5-15, and 8.3% in symptomatic mild OSA. The 'normal' sleeper is not breathing perfectly. They are breathing less imperfectly than the diagnosable patient.

The Pediatric Dose-Response

The pediatric literature provides the most compelling evidence that respiratory effects operate below any diagnostic threshold. In a community-based sample of 1,010 children aged 5-7 years, neurocognitive performance declined in a linear, dose-response relationship with sleep-disordered breathing severity. Critically, this decline does not begin at the AHI cutoff. Snoring status alone, entirely below diagnostic thresholds, independently predicts behavioral and cognitive deficits after adjusting for AHI severity and baseline covariates. Furthermore, longitudinal data shows that even when pediatric SDB resolves (reaching an obstructive AHI below 1), cognitive and behavioral deficits persist, suggesting that sub-threshold respiratory events inflict durable neurodevelopmental alterations.

Two Sleepers, Same Hours, Different Outcomes

The continuum model predicts a phenomenon that standard sleep measurement cannot explain: two individuals with identical sleep staging, identical continuity metrics, and identical total sleep time can have dramatically different restorative outcomes. The variable that distinguishes them is sub-architectural sleep quality, governed by the respiratory control variables described above.

Person A has low loop gain, a wide CO₂ reserve, robust cerebrovascular reactivity, strong upper airway dilator tone, and high parasympathetic dominance during sleep. Their breathing is smooth and stable throughout the night. Cerebral blood flow is well-maintained. Glymphatic clearance operates efficiently. Tissue oxygenation through the Bohr effect is optimal. Every hour of sleep delivers maximum restorative value.

Person B has higher (but still 'normal') loop gain, a narrower CO₂ reserve, blunted CVR, moderate inspiratory flow limitation, and residual sympathetic tone from chronic stress-driven overbreathing. Their PSG looks the same as Person A's: same stages, same continuity, same total time. But their breathing during those stages includes sub-threshold oscillations, flow limitation, micro-arousals too brief to score, and transient hypocapnic excursions that repeatedly narrow the operating margin without ever crossing the apneic threshold. Cerebral blood flow fluctuates. Glymphatic clearance is disrupted. Tissue oxygenation is intermittently impaired. Every hour of sleep delivers less restorative value than Person A's.

Person A wakes after 6 hours feeling fully restored. Person B wakes after 8 hours feeling unrefreshed. Standard polysomnography cannot distinguish them. The AHI is identical (below 5 for both). The sleep staging is comparable. The difference is entirely in the respiratory control variables that determine how much restorative work each minute of sleep actually accomplishes.

The Concept of Respiratory Fitness

Cardiovascular fitness is a universally accepted concept. An individual's VO₂ max, resting heart rate, stroke volume, and cardiac output are understood as trainable traits that exist on a continuum from severely deconditioned to elite. Nobody argues that cardiovascular fitness only matters for people with heart disease. The concept applies to the entire population, and training produces measurable improvement at every point on the spectrum.

This paper proposes that respiratory fitness should be understood in the same way. An individual's loop gain, CO₂ tolerance, chemoreceptor sensitivity, cerebrovascular reactivity, inspiratory muscle strength, upper airway dilator tone, and habitual breathing pattern are trainable traits that influence sleep quality the way cardiovascular conditioning influences exercise performance. Low respiratory fitness produces poor sleep. High respiratory fitness produces efficient sleep. The relationship is continuous, not threshold-dependent. Training produces improvement at every point on the spectrum, not only in clinical populations.

Formally defined: respiratory fitness is the integrated capacity of the ventilatory control system, upper airway musculature, respiratory pump, cerebrovascular buffer, and autonomic nervous system to maintain stable, efficient, nasal-dominant breathing across sleep and wake states. It is measurable through the metrics described in section “Measurable Outcomes for Respiratory Fitness”, modifiable through the training stack described in section “From Treatment to Training: The Paradigm Shift”, and predictive of sleep quality independently of sleep duration or staging.

The components of respiratory fitness map directly onto the five variables identified in section “The Respiratory Variables That Govern Sleep Quality”. Each is measurable, each is modifiable, and each has evidence supporting the relationship between its optimization and improved sleep outcomes.

Evidence That Respiratory Variables Are Trainable

CO₂ Tolerance Training

The most direct evidence that chemoreceptor sensitivity is trainable comes from extreme physiology and breath-hold research. Elite freedivers develop profoundly blunted chemoreflex responses through chronic exposure to hypercapnia and hypoxia. Their physiological adaptations include delayed onset of involuntary diaphragmatic contractions, enhanced blood buffering capacity through increased bicarbonate retention, and maintained cerebral blood flow at CO₂ levels that would cause syncope in untrained individuals. These adaptations represent a direct, measurable reduction in controller gain.

In clinical populations, structured breath-hold training (Buteyko methodology) has produced measurable improvements in polysomnographic sleep architecture. In a controlled PSG study, subjects undergoing respiratory retraining showed total sleep time increasing from 356.2 to 383.2 minutes and overall sleep efficiency improving from 82.8% to 87.3%, with decreased REM latency. The mechanism proposed is not necessarily a permanent change in resting blood gases (which was not observed) but rather a modulation of autonomic tone: dampened sympathetic hyperarousal, reduced frequency of ventilatory overshoots, and enhanced parasympathetic dominance facilitating stable, uninterrupted sleep.

The altitude physiology literature provides a natural experiment. High-altitude-adapted Sherpa populations exhibit genetically blunted ventilatory responses to CO₂, producing inherently low controller gain. This protects their CO₂ reserve at altitude, shielding them from the periodic breathing that devastates sleep in unacclimatized lowlanders. The Sherpa phenotype represents the optimal end of the respiratory fitness continuum: low loop gain, wide CO₂ reserve, stable breathing under extreme conditions, preserved sleep architecture.

Inspiratory Muscle Training

Inspiratory muscle training (IMT) strengthens the mechanical components of respiratory fitness. In systematic reviews and randomized controlled trials, IMT increases maximum inspiratory pressure (MIP), reduces AHI, improves Pittsburgh Sleep Quality Index scores, and decreases excessive daytime sleepiness in OSA populations. The proposed mechanism is a reduction in plant gain: stronger airway dilators make the pharynx less susceptible to collapse during negative-pressure inspiratory cycles.

Critically, the benefits extend to non-clinical populations. A prospective randomized controlled trial of 67 healthy adults using combined inspiratory and expiratory respiratory muscle training (cRMT) for five weeks produced a significant increase in overnight heart rate variability, specifically in the high-frequency power band reflecting parasympathetic activity. While macroscopic sleep staging did not change, the autonomic environment of sleep was measurably improved. This finding demonstrates that respiratory training can enhance the quality of sleep in individuals with no diagnosable sleep disorder, operating through a mechanism that standard sleep metrics do not capture.

Nasal Breathing

The transition from habitual oral to nasal breathing is the simplest and most accessible respiratory fitness intervention. Oral breathing during sleep compromises the pharyngeal airspace through mandibular drop, tongue base retraction, and loss of physiological PEEP. Nasal breathing provides mechanical airway stabilization, entrains nitric oxide for improved ventilation-perfusion matching, increases airway resistance to physiologically slow the respiratory rate, and promotes diaphragmatic rather than thoracic breathing patterns.

In controlled clinical studies, structured unilateral nasal breathing protocols significantly improved all seven component scores of the Pittsburgh Sleep Quality Index over four weeks. The mechanism is hypothesized to operate through enhanced parasympathetic dominance during NREM sleep, dampening the sympathetic hyperarousal that triggers flow limitation and sub-threshold arousals. In athletic populations, nasal-only breathing during exercise produces lower heart rates at equivalent workloads and faster post-exercise tissue resaturation, indicating superior cardiovascular efficiency.

The Respiratory Fitness Continuum

Integrating the evidence from sections “The Respiratory Variables That Govern Sleep Quality” through 6, the following continuum emerges:

The critical insight is that movement along this continuum is possible in both directions. Chronic stress, sedentary lifestyle, habitual mouth breathing, and anxiety can push an individual from baseline healthy toward sub-optimal or deconditioned over months to years. Targeted respiratory training, including CO₂ tolerance work, inspiratory muscle training, nasal breathing discipline, and autonomic regulation practices, can push an individual from sub-optimal toward optimized on a similar timescale.

The Sub-Optimal Majority: A Hidden Epidemic

The population stranded between 'diagnosable' and 'optimized' on this continuum is enormous. Epidemiological data shows that 24-26% of adult men and 9-28% of adult women have an AHI of 5 or above, but only 3-7.5% meet the combined criteria of elevated AHI plus excessive daytime sleepiness required for the classic actionable OSA phenotype. The remainder have measurable respiratory disruption but are medically dismissed as asymptomatic or sub-clinical.

Below the diagnostic threshold, an even larger population suffers from poor sleep quality without explanation. Individuals with insomnia symptoms who do not meet criteria for full syndromal insomnia still cost the economy an average of $1,431 per person annually, more than triple the $421 baseline cost of good sleepers. In the United States alone, inadequate sleep costs an estimated $411 billion annually, representing 2.28% of GDP, driven primarily by presenteeism: employees who are physically present but cognitively degraded from unrefreshing sleep.

The respiratory fitness framework offers an explanation for this population's symptoms and, more importantly, an intervention pathway. These individuals do not have a sleep disorder. They have low respiratory fitness. Their loop gain is at the higher end of normal. Their CO₂ reserve is adequate but not robust. Their habitual breathing patterns reflect years of stress-driven overventilation. Their autonomic balance is tilted toward sympathetic dominance. None of these variables would flag on a clinical evaluation, but collectively they constrain sleep efficiency enough to produce the chronic unrefreshing sleep that drives them to seek help.

The current system tells these individuals their sleep is normal. The respiratory fitness framework tells them their sleep is undertrained, and gives them a specific, measurable path to improve it.

From Treatment to Training: The Paradigm Shift

The prevailing model of sleep medicine is disease-focused. It identifies pathology, applies treatment (CPAP, oral appliances, surgery, pharmacology), and measures success by reduction of scoreable events. This model has been enormously successful for severe, diagnosable sleep apnea. It has nothing to offer the sub-optimal majority because their condition does not meet the threshold for pathology.

The respiratory fitness model is performance-focused. It treats the respiratory variables governing sleep quality as trainable traits that can be systematically improved through targeted interventions. The interventions are not treatments for a disease. They are training protocols for a physiological system, analogous to cardiovascular training programs that improve cardiac performance across the entire population spectrum.

The respiratory fitness training stack, from most accessible to most specialized:

Tier 1: Nasal breathing. Transition from habitual oral to nasal breathing during both sleep and exercise. This is the simplest intervention with the broadest applicability. It provides mechanical airway stabilization, entrains nitric oxide, increases airway resistance to slow respiratory rate, and promotes diaphragmatic breathing. Lip taping during sleep may be appropriate for some individuals.

Tier 2: Autonomic regulation. Structured slow-breathing protocols (4-7-8 breathing, extended-exhale techniques, resonance frequency breathing) that shift autonomic balance toward parasympathetic dominance. These reduce sympathetic hyperarousal, improve HRV, and create the neurochemical conditions for stable sleep onset and consolidation.

Tier 3: CO₂ tolerance training. Progressive breath-hold work that gradually increases the chemoreceptor tolerance for elevated CO₂, widening the CO₂ reserve and reducing controller gain. This is the most direct intervention for loop gain reduction but also the most demanding, as it requires tolerating the discomfort of air hunger. The 'Geno dilemma': most patients will not sustain this training voluntarily, which is precisely why device-based passive CO₂ conditioning may be necessary to make this intervention scalable.

Tier 4: Inspiratory muscle training. Resistance-based strengthening of the diaphragm and upper airway dilator muscles, increasing MIP and reducing the mechanical vulnerability of the airway to collapse under negative inspiratory pressure.

Tier 5: Device-based CO₂ conditioning. Closed-loop, capnography-guided systems that deliver titrated CO₂ exposure during sleep or pre-sleep to passively condition the chemoreceptors without requiring conscious effort or discomfort tolerance. This represents the scalable, compliance-friendly version of Tier 3 and is the frontier of respiratory fitness technology. Enhanced Expiratory Rebreathing Space (EERS), developed by Thomas et al. at Beth Israel Deaconess, demonstrates the acute principle. The translational step is developing EERS-style exposure into a chronic conditioning platform.

Measurable Outcomes for Respiratory Fitness

If respiratory fitness is a trainable trait, it requires measurable outcomes to track progress. The following metrics are proposed as a respiratory fitness assessment framework, ordered from most accessible to most specialized:

N-of-1 Case Illustration

The framework is illustrated by a documented case of a middle-aged male with treated OSA (AHI consistently below 5 on CPAP) who experienced years of severe residual sleep fragmentation: 12-15 awakenings per night, 9+ hours of non-restorative sleep, daily anxiety at 7/10, caffeine and chemical sleep aid dependence, and progressive functional narrowing. Standard sleep medicine had no explanation or intervention because the AHI was controlled.

Over approximately twelve months of systematic respiratory fitness training, primarily progressive CO₂ tolerance work (breath-hold duration progressing from approximately 10 seconds to 45 seconds) combined with strict nasal-only breathing during all sleep and exercise, the following changes occurred: awakenings fell from 12-15 per night to 1-3; total sleep need dropped from 9+ hours to approximately 6 hours of fully restorative sleep; anxiety resolved to 0/10; all chemical sleep aids and caffeine dependence resolved; full behavioral function returned. Sixteen consecutive days of post-recovery tracking documented zero instances of adenosine-driven daytime fatigue, with average waking function extending 16+ hours daily.

This case illustrates the continuum model precisely. The individual was never 'cured' of a disease. The disease (obstructive sleep apnea) was already treated. What changed was respiratory fitness: CO₂ tolerance, loop gain, ventilatory stability, and autonomic balance were trained from severely deconditioned to optimized. The sleep improvement was not a treatment effect. It was a training effect. The mechanism is detailed in a companion paper: 'Chronic Behavioral Hyperventilation-Induced Sleep Fragmentation: An N-of-1 Recovery Account and Mechanistic Model.'

Implications and Future Directions

If the respiratory fitness framework is correct, several implications follow:

Sleep medicine should expand beyond pathology treatment to include respiratory fitness assessment and training as a standard component of sleep quality optimization for all patients presenting with unrefreshing sleep, regardless of AHI.

The development of passive, device-based CO₂ conditioning systems (building on the EERS proof of concept) should be pursued as a scalable intervention for the sub-optimal majority who cannot or will not perform voluntary breath-hold training.

Respiratory fitness metrics (breath-hold duration, resting respiratory rate, HRV, habitual breathing route) should be incorporated into consumer sleep tracking platforms alongside the sleep staging and heart rate data they already capture.

Research protocols studying sleep quality should routinely include respiratory control variables (end-tidal CO₂, loop gain estimates, capnography, MIP) as potential mediators or predictors, rather than relying exclusively on AHI and sleep staging.

The enormous population of poor sleepers with no diagnosable disorder, currently told their sleep is 'normal,' should be recognized as individuals with low respiratory fitness who would benefit from targeted training rather than pharmacological sedation or sleep hygiene advice.

The economic case for respiratory fitness training is substantial. If even a fraction of the $411 billion annual US cost of inadequate sleep could be addressed through scalable respiratory interventions rather than medical treatment, the return on investment in respiratory fitness research and device development would be extraordinary.

Proposed Study Design

To test the respiratory fitness framework empirically, the following study design is proposed:

Population: Adults aged 25-55 reporting chronically unrefreshing sleep or excessive sleep need (8+ hours with residual fatigue) who do not meet diagnostic criteria for OSA (AHI below 5), insomnia, or other formal sleep disorders. This is the sub-optimal majority: individuals with no diagnosis, no treatment pathway, and no explanation for their symptoms.

Baseline assessment: End-tidal CO₂, resting respiratory rate, maximum inspiratory pressure (MIP), breath-hold duration (Control Pause), habitual breathing route (oral vs. nasal, assessed during sleep via video or sensor), heart rate variability (HF power), Pittsburgh Sleep Quality Index, and full polysomnography with concurrent capnography and loop gain estimation.

Intervention (12 weeks): A tiered respiratory fitness protocol: (1) nasal breathing transition including lip taping during sleep; (2) daily structured slow-breathing practice (4-7-8 or resonance frequency breathing, 10 minutes); (3) progressive CO₂ tolerance training through breath-hold work, titrated to individual tolerance, 3-5 sessions per week; (4) inspiratory muscle training using a pressure-threshold device, 30 breaths twice daily at 50-75% of MIP.

Control group: Sleep hygiene education only (the current standard of care for this population).

Outcome measures (12 weeks): Repeat of all baseline measures. Primary outcomes: change in sleep efficiency on PSG, change in total arousal index, change in nocturnal PaCO₂ variability on capnography. Secondary outcomes: change in PSQI, change in breath-hold duration, change in HRV, change in subjective sleep need, change in daytime fatigue measures.

Predicted findings: The respiratory fitness group should show reduced nocturnal CO₂ variability, reduced arousal index, improved sleep efficiency, improved HRV, increased breath-hold duration, and improved subjective sleep quality relative to the sleep hygiene control group. If the framework is correct, the magnitude of improvement should correlate with the magnitude of change in respiratory control variables (particularly loop gain and CO₂ reserve) rather than with baseline sleep severity.

Limitations

This paper synthesizes evidence across multiple domains (respiratory physiology, sleep medicine, altitude physiology, exercise science, epidemiology) to propose a unified framework. The individual components are supported by published evidence. The integration into a single 'respiratory fitness' construct is the novel contribution and has not been validated as such.

The relationship between respiratory training and sleep quality improvement in non-clinical populations remains understudied. The strongest evidence comes from clinical populations (OSA patients) and extreme physiology (freedivers, altitude). Controlled trials of respiratory fitness training in the sub-optimal majority are needed to establish the breadth of applicability.

The 'respiratory fitness' concept intentionally parallels 'cardiovascular fitness' as a framing device. The analogy is useful but imperfect. Cardiovascular fitness has decades of validated measurement tools, dose-response data, and population-level studies. Respiratory fitness, as proposed here, requires similar investment to achieve comparable scientific maturity.

The N-of-1 case described in section “N-of-1 Case Illustration” illustrates the framework but does not constitute evidence for it. Controlled studies with appropriate sample sizes, blinding, and objective respiratory and sleep outcome measures are necessary to determine whether the respiratory fitness training stack produces consistent, reproducible improvements across the population spectrum.

Sources and related reading

This manuscript does not contain a standalone numbered bibliography. Follow the evidence library and the related core manuscripts for source-linked background.