Researchers at the University of Waterloo recruited twenty-two healthy young adults — eleven women and eleven men — who exercised more than three times per week. Each participant completed a single laboratory session involving three very different exercise tasks: rhythmic handgrip, which activates only the small muscles of the forearm; single-leg extension, which engages the quadriceps and surrounding muscles of one limb; and cycling, a whole-body activity recruiting large muscle groups in both legs. The team, led by Armando J. Aguero, Connor J. Doherty and Paolo B. Dominelli, hypothesized that chemosensitivity would climb in step with the amount of active muscle, so that cycling would produce the largest boost and handgrip the smallest.

To measure peripheral hypercapnic chemosensitivity, the researchers used a clever transient stimulus. On four occasions during each resting and exercising condition, participants unknowingly received two breaths of a gas mixture containing 10 percent carbon dioxide, delivered through a silent manual valve switched by investigators watching real-time ventilatory flow. Because the stimulus lasted only two breaths, it was far too brief to engage the brain’s central chemoreceptors, which respond to slower, sustained changes in carbon dioxide. Instead, the response was attributed to the peripheral chemoreceptors, chiefly the carotid bodies. The sensitivity was then calculated as the change in inspired ventilation divided by the change in end-tidal carbon dioxide pressure, expressed in liters per minute per millimeter of mercury.

The exercise intensities were deliberately kept low. Handgrip was performed against a 2.5-kilogram load at thirty contractions per minute with a one-second-on, one-second-off rhythm. Single-leg extension pushed against a constant five watts of resistance on a custom ergometer at the same cadence. Cycling was performed at forty watts at a self-selected cadence of roughly sixty to eighty revolutions per minute. These modest workloads were chosen to ensure that the metabolic byproducts of exercise would be minimal, reducing the contribution of the muscle metaboreflex — the reflex triggered by metabolite accumulation in working muscle — and allowing the team to focus on neural mechanisms such as central command and mechanoreceptor feedback.

The results confirmed half of the hypothesis. All three exercise modalities produced a significant increase in peripheral hypercapnic chemosensitivity from rest to exercise. Handgrip raised sensitivity by about 11.9 percent, single-leg extension by roughly 52.4 percent, and cycling by approximately 49.2 percent. Heart rate, ventilation, tidal volume and the ventilatory response to the carbon dioxide bolus all rose during the active conditions, with the largest changes seen in the leg exercises. But the critical comparison between single-leg extension and cycling revealed the surprise: despite cycling recruiting far more muscle mass, the two exercises produced statistically indistinguishable increases in chemosensitivity, both in absolute terms and as a percentage change from baseline.

This finding challenges a straightforward assumption about how the body scales its ventilatory response. The researchers had reasoned that larger active muscle mass would demand greater central command — the feedforward signal from the brain’s motor and autonomic centers that drives anticipatory increases in heart rate and ventilation before and during exercise. Central command is known to increase in proportion to both exercise intensity and the amount of muscle recruited, progressing from isolated small muscle groups to combined limb exercise. If central command were the sole driver of chemosensitization, cycling should have outpaced single-leg extension. It did not.

The study’s authors suggest instead that the key factor is the transition from rest to exercise itself, rather than the magnitude of the effort. This interpretation is consistent with their earlier work showing that chemosensitivity rises when cycling begins but plateaus with further increases in workload. Together, the two findings paint a picture in which the act of initiating movement — engaging central command and muscle mechanoreceptors — flips a switch that sensitizes the carotid bodies, but adding more muscle or more intensity does not turn that switch further. The similar responses during single-leg extension and cycling imply that once a threshold of neural activation is crossed, recruiting additional muscle mass adds little to the chemosensory gain.

Mechanoreceptor feedback may also play a role. Type III muscle afferents respond to mechanical stretch and contraction and can augment sympathetic outflow and ventilatory drive early in exercise. Because single-leg extension and cycling stimulate more mechanoreceptors than handgrip, the intermediate rise in chemosensitivity seen in the leg exercises is most consistent with concurrent signaling from both central command and mechanoreceptor feedback. The authors caution, however, that voluntary exercise inherently involves both mechanisms, and their design cannot fully separate the contributions of each. Type IV chemoreceptive afferents, which respond to metabolic stress, are unlikely culprits given the low exercise intensities used and prior evidence that metaboreceptor feedback does not increase chemosensitivity beyond the onset of exercise.

Several methodological considerations shaped the interpretation. Resting chemosensitivity was higher when participants were seated on the cycle ergometer than during handgrip, which the authors attribute to an anticipatory response driven by central command before exercise begins — a phenomenon they had observed previously. Pre-stimulus end-tidal carbon dioxide also differed across conditions: it rose during the leg exercises, creating a mildly hypercapnic environment that would amplify the ventilatory response, but fell during handgrip, which should have blunted the response. Notably, handgrip still produced a significant increase in chemosensitivity despite this hypocapnic baseline, demonstrating that the exercise-related sensitization was robust enough to overcome an attenuating condition. The authors also note that end-tidal carbon dioxide serves as a surrogate for arterial carbon dioxide, a substitution that is reasonable at low exercise intensities where the gradient between the two is minimal.

The broader implications reach into exercise physiology, respiratory medicine and perhaps athletic performance. Understanding how the carotid bodies are sensitized during movement could inform approaches to conditions in which ventilatory control is disordered, and the finding that chemosensitization does not scale with muscle mass simplifies the search for the underlying neural signal. The authors propose that future studies use electrically evoked, non-volitional muscle contractions to activate muscle without engaging central command, or block muscle afferents with intrathecal fentanyl while preserving motor drive, to disentangle the two candidate mechanisms. For now, the message is clear: when it comes to sharpening the body’s carbon dioxide sensors, starting to move matters far more than how much muscle you move with.

Subject of Research: Effect of active muscle mass on peripheral hypercapnic chemosensitivity during exercise in healthy adults

Article Title: Greater muscle mass does not affect peripheral hypercapnic chemosensitivity during exercise in healthy adults

Article References: Aguero, A. J., Doherty, C. J., & Dominelli, P. B. (2026). Greater muscle mass does not affect peripheral hypercapnic chemosensitivity during exercise in healthy adults. Physiological Reports, 14(17), Article e71102. https://doi.org/10.14814/phy2.71102

Image Credits: AI Generated

DOI: 10.14814/phy2.71102

Keywords: peripheral hypercapnic chemosensitivity, carotid bodies, exercise physiology, central command, mechanoreceptor feedback, ventilatory control, carbon dioxide sensitivity, rhythmic handgrip, cycling, muscle afferents, Physiological Reports, respiratory physiology

The study, led by Azarbarzin and colleagues and published in the Journal of Clinical Sleep Medicine, drew on 7,530 participants from two well-established cohort studies, the Sleep Heart Health Study and the Osteoporotic Fractures in Men Study, known as MrOS. Over nearly nine years of follow-up, 2,258 major adverse cardiovascular events occurred in this population. The investigators computed an oxygen overshoot burden for each participant: the cumulative area under the oxygen saturation curve that lies above an individual’s stable-sleep baseline, normalized for total sleep time. This is a deliberately individualized measure. Rather than comparing every patient against a fixed cutoff, the analysis anchors the signal to each person’s own resting saturation during uneventful sleep, so that even small relative rises above the personal baseline accumulate into a meaningful summary of nightly physiological stress.

The headline result is striking. Among participants with central sleep apnea and high overshoot burden, 54.6 percent experienced a major adverse cardiovascular event during follow-up, compared with 25.6 percent of control participants, corresponding to an adjusted hazard ratio of 1.45. Crucially, central sleep apnea with low overshoot burden carried no elevated risk relative to controls, and no corresponding association appeared in obstructive sleep apnea. The relationship survived statistical adjustment for the apnea-hypopnea index, the central apnea index, hypoxic burden, and, in the MrOS cohort, an estimate of loop gain, the control-system parameter that describes how vigorously breathing responds to disturbances in blood gases. That specificity is intriguing: it suggests the overshoot signal is not merely a proxy for how often someone stops breathing, but may capture something distinct about the instability of their ventilatory control.

The authors and commentators have proposed an intuitive biological explanation rooted in oxidative stress. The cycle of desaturation and reoxygenation that characterizes sleep apnea resembles ischemia-reperfusion injury, and repeated swings are thought to generate reactive oxygen species, inflammation, and vascular damage. Hypoxic burden, which integrates the depth and duration of event-related desaturation, already predicts cardiovascular outcomes better than a simple count of respiratory events, lending weight to the idea that the shape of the oxygen trace carries mechanistic information. Extending that logic above the baseline line is tempting: if falling oxygen is harmful, perhaps overshooting oxygen is harmful too, and the cumulative area of overshoot might quantify a dose of oxidative injury delivered night after night.

Yet a careful reading of the physiology counsels caution before accepting that interpretation. Oxygen overshoot as measured here is a relative oximetry signal, not demonstrated hyperoxia. A rise in peripheral oxygen saturation above a person’s stable-sleep baseline does not establish an elevated arterial partial pressure of oxygen, nor does it demonstrate increased tissue oxygen exposure. The study provides no mechanistic pathway, no oxidative biomarker measurements, and no bench or animal evidence showing that this specific above-baseline signal causes oxidative injury. Evidence from the intermittent hypoxia and reoxygenation literature cannot simply be transferred to this different signal. Indeed, previous work in obstructive sleep apnea found that greater post-event saturation overshoot was associated with lower nocturnal glucose, a pattern that argues against harm and even hints at benefit. Treating the oxygen rise itself as the causal exposure is, for now, premature.

A more plausible reading is that the oximetric overshoot is the visible tail of a much larger ventilatory and autonomic response. When an apnea terminates, the accumulated carbon dioxide and chemoreflex drive produce a vigorous recovery breath; the resulting hypocapnia is a direct signature of unstable ventilatory control, which is precisely the physiology that generates central sleep apnea and Cheyne-Stokes respiration in the first place. Human studies of apnea have documented marked sympathetic nerve activation and blood pressure surges around the termination of each event, followed by vagal modulation tied to lung inflation during the recovery phase. These autonomic oscillations, repeated hundreds of times a night, offer a credible route to myocardial infarction, arrhythmia, and stroke that does not require a modest rise in saturation to be intrinsically toxic. On this view, overshoot burden is a marker of the force of each recovery, not a poison in its own right.

This interpretation also exposes an important analytical gap in the new study. Adjusting for event frequency and hypoxic burden does not establish that overshoot is independent of the severity of each respiratory event. A longer or more severe apnea accumulates more hypercapnia, more chemoreflex stimulation, more arousal-related and sympathetic activation, and a longer loss of the vagal restraint normally provided by lung inflation. Those stimuli can then generate a larger recovery breath and a larger oxygen overshoot, and desaturation depth alone does not fully represent them. Consistent with this, work in obstructive sleep apnea has shown that ventilatory burden, a measure of the effort expended against collapsed airways, predicts cardiovascular outcomes and explains much of the variation in hypoxic burden. Oxygen overshoot may play an analogous role in central sleep apnea, summarizing a hidden physiological load that conventional indices leave unmeasured, without being the injurious agent itself.

Residual cardiac confounding remains another live possibility. In older community cohorts, central sleep apnea and Cheyne-Stokes respiration often reflect underlying cardiac dysfunction, elevated left-sided filling pressures, and prolonged circulation time, rather than mediating the cardiovascular consequences of those conditions. Heart failure identified through self-report and clinical records may miss subclinical disease, so some of the apparent association between overshoot burden and events could reflect undiagnosed cardiac impairment that both destabilizes breathing control and drives outcomes. Adjustment for loop gain in one cohort helps address the ventilatory-instability pathway, but detailed cardiac phenotyping, with objective measures of structure and function, is still needed before the direction of the arrow can be declared with confidence.

Two further cautions temper clinical translation. The central sleep apnea subgroup comprised only 303 participants, so confidence intervals around the headline event proportions deserve as much attention as the proportions themselves, and the findings require replication in larger and more diverse samples. The demonstration that a subject-specific baseline outperformed a fixed threshold of saturation at or above 96 percent confirms that the result depends on small relative differences in the oximetry signal, which places a premium on signal quality and measurement precision. The cohorts were 88 percent White, and known differential pulse-oximeter error by skin pigmentation is relevant to any saturation-based metric, although its effect on an above-baseline area calculation is currently unknown and warrants direct study.

The path forward is clear enough. The next study should measure event duration, airflow, respiratory effort, carbon dioxide, arousals, and event-level autonomic responses alongside oximetry, with detailed cardiac phenotyping and representative recruitment, and then test whether oxygen overshoot adds prognostic information once those features are accounted for. Until that work is done, oxygen overshoot burden should be regarded as a promising risk marker, a way of reading the recovery half of the apnea cycle that standard metrics ignore, rather than a demonstrated oxidative mechanism or a validated treatment target. Even so, the study is a reminder that the familiar oximetry trace still holds unexploited information, and that the line above the dips may matter as much as the dips themselves.

Subject of Research: Oxygen overshoot burden measured on overnight oximetry traces in central sleep apnea and its association with major adverse cardiovascular events.

Article Title: Above the line: oxygen overshoot in the oximetry trace

Article References: Manuel, A. R. G. (2026). Above the line: oxygen overshoot in the oximetry trace. Journal of Clinical Sleep Medicine, 22(1), Article 169. https://doi.org/10.1007/s44470-026-00183-8

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00183-8

Keywords: central sleep apnea, oxygen overshoot, pulse oximetry, Cheyne-Stokes respiration, cardiovascular risk, hypoxic burden, ventilatory control, autonomic surges, oxidative stress, Sleep Heart Health Study, loop gain, sleep apnea