REM Sleep Holds the First Clues That Children With Muscle Disease Need Breathing Support

For children living with neuromuscular diseases such as Duchenne muscular dystrophy and spinal muscular atrophy, the quiet hours of the night can be the most dangerous part of the day. As respiratory muscles weaken, breathing during sleep deteriorates long before any daytime symptoms appear, and the earliest warning signs often surface during rapid eye movement, or REM, sleep, when the body’s muscles are naturally paralyzed. A new multicenter study from Canada, published in the Journal of Clinical Sleep Medicine, suggests that the newest clinical guidelines for deciding when to start non-invasive ventilation are better than older, widely used criteria at catching these children before their sleep-disordered breathing becomes advanced.

The research, led by Vanessa Campes Dannenberg and Maria L. Castro-Codesal of the University of Alberta, together with colleagues at the Hospital for Sick Children in Toronto and the Alberta Children’s Hospital in Calgary, examined the medical records of 115 children and young people with neuromuscular disorders who were recommended for home non-invasive ventilation, or NIV, between 2010 and 2023. Most of the participants were male, with a median age of 11.8 years at the time of ventilation initiation. The most common diagnoses were Duchenne muscular dystrophy, accounting for 39 percent of the cohort, congenital muscular dystrophies at 19 percent, and spinal muscular atrophy at 14 percent. Nearly every child in the study carried at least one additional chronic condition, with a median of four comorbidities per patient, underscoring how medically complex this population is.

The central question was deceptively simple: when three different sets of polysomnography-based criteria are applied to the same overnight sleep study, do they identify the same children as needing breathing support? The team compared the 2023 guidelines from the American College of Chest Physicians, known as CHEST, against the traditional American Academy of Sleep Medicine, or AASM, definitions of moderate-to-severe obstructive sleep apnea and nocturnal hypoventilation, and against a new set of REM-related criteria proposed by the authors themselves. The CHEST criteria, drawn from a European Respiratory Society statement on pediatric long-term ventilation, flag children whose end-tidal or transcutaneous carbon dioxide exceeds 50 mmHg for at least 2 percent of total sleep time, or rises by 10 mmHg above their waking baseline, or whose oxygen saturation drops below 90 percent for at least 2 percent of sleep time or five continuous minutes.

Those thresholds matter because they are far more sensitive than the older AASM definitions, which require carbon dioxide to exceed 50 mmHg for a full 25 percent of total sleep time before labeling a child as hypoventilating. The results were striking. Only 9 percent of the children met the AASM-based hypoventilation criterion, while the CHEST guidelines identified 25 percent, and the REM-related criteria caught 12 percent. For nocturnal hypoxemia, for which the AASM framework offers no initiation criterion at all, the CHEST guidelines identified 22 percent of the cohort and the REM-related criteria 15 percent. In total, 71 percent of the children met at least one NIV initiation criterion across the three criteria sets, but a meaningful minority were visible only through the newer lenses.

Agreement between the criteria sets, measured with kappa statistics, told a nuanced story. Overall agreement between the CHEST and REM-related criteria was moderate, with a kappa of 0.71, and agreement on nocturnal hypoxemia between those two sets was similarly moderate at 0.69. The strongest agreement, a kappa of 0.86, emerged between the CHEST and AASM criteria on the apnea-hypopnea index, which is unsurprising given that both frameworks use similar thresholds of five or more respiratory events per hour. The divergence, in other words, was not about counting apneas and hypopneas. It was about the slow, silent buildup of carbon dioxide and the dips in oxygen that occur without any discrete respiratory event, particularly during the muscle atonia of REM sleep.

That physiology is the heart of the study’s argument. In children with neuromuscular weakness, breathing during REM sleep depends heavily on the diaphragm, because the intercostal and upper airway muscles lose their tone in that stage. When the diaphragm is also weak, carbon dioxide accumulates and oxygen levels fall, often without producing a single scored apnea. As the disease progresses, these abnormalities spread into non-REM sleep and eventually into wakefulness, culminating in chronic respiratory failure. The authors’ proposed REM-related criteria were designed to capture this earliest phase, and the analysis showed that an additional 11 percent of children met REM-related indications that the AASM-based criteria missed entirely. Thirteen percent of the cohort, twelve children, met no AASM-based criteria at all but satisfied at least one CHEST and one REM-related criterion.

The study also looked at what happened after ventilation began. Ninety-six percent of the children, 111 in total, successfully started NIV, and 73 percent of them sustained the therapy for six to twelve months, the window the researchers chose to avoid the acclimatization period when adherence is typically unstable. Adherence was respectable: children used the machine for a median of six hours per night, and on a median of 60 percent of nights they managed more than four hours of use. Clinician-documented reports from routine visits described improvements in general health in 34 percent of children, better sleep quality in 36 percent, and improved breathing during sleep in 25 percent. Most children avoided emergency department visits, ward admissions, and intensive care during the follow-up period.

Notably, the study found no significant differences in lung function before and after NIV initiation in any of the criteria groups, and no differences in adherence or clinical outcomes between groups. The authors attribute this largely to the substantial overlap among the three groups, since most children met the apnea-hypopnea thresholds common to all criteria sets. They also caution that NIV does not reverse the decline in lung function driven by the underlying disease; at best, it slows the trajectory, and changes are influenced by disease-modifying therapies and the natural history of each disorder. Only about half of the cohort could perform lung function testing that met American Thoracic Society technical standards, further limiting those comparisons.

The findings carry real weight for clinical practice. Earlier work in children with Duchenne muscular dystrophy had already suggested that disease-specific criteria could prompt ventilation for an additional 22 percent of children compared with standard AASM definitions. This study extends that insight to a broader neuromuscular population and adds a physiological rationale: many children show REM-related hypoventilation or hypoxemia with a normal apnea-hypopnea index, and the CHEST guidelines, by lowering the carbon dioxide threshold and adding an explicit hypoxemia criterion, are built to detect exactly that pattern. This matters especially for conditions such as myotonic dystrophy, in which discrete respiratory events are less common but carbon dioxide retention can still be dangerous.

The researchers acknowledge the limits of their retrospective design, their reliance on chart documentation, and the overlap that prevented subgroup comparisons. They also note that future studies may be difficult to conduct, because withholding ventilation until advanced hypoventilation develops would now be ethically questionable. Still, their conclusion is clear and forward-looking: guidelines for children with neuromuscular disease should integrate disease-specific and sleep-stage-specific polysomnographic criteria, watching REM sleep closely for the first signs of failure. Caught early, supported by a mask and a machine at home, these children may avoid the steepest part of the respiratory decline that has long defined their prognosis.

Subject of Research: Polysomnographic criteria for initiating non-invasive ventilation in children with neuromuscular diseases

Article Title: Establishing polysomnographic criteria for initiation of non-invasive ventilation in children with neuromuscular diseases

Article References: Dannenberg, V. C., Amin, R., Anagnostopoulos, G., Adeleye, A., Bendiak, G. N., Kolski, H., & Castro-Codesal, M. L. (2026). Establishing polysomnographic criteria for initiation of non-invasive ventilation in children with neuromuscular diseases. Journal of Clinical Sleep Medicine, 22(1), Article 117. https://doi.org/10.1007/s44470-026-00102-x

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00102-x

Keywords: neuromuscular disease, non-invasive ventilation, polysomnography, REM sleep, sleep-disordered breathing, nocturnal hypoventilation, Duchenne muscular dystrophy, spinal muscular atrophy, pediatric sleep medicine, CHEST guidelines, AASM criteria, nocturnal hypoxemia