The letter, written by Christoph G.U. Riese and Ulrich Koehler of the Sleep Medicine Center at Philipps-Universität Marburg in Germany, responds to a larger investigation known as the MAGNETO study. That study, led by Junco and colleagues, made an unsettling discovery: patients with treated sleep apnea who still reported excessive daytime sleepiness also showed measurable cognitive deficits, including impairments in attention and psychomotor vigilance, the capacity to react quickly and consistently to incoming signals. Even after the researchers statistically adjusted for sex, the delay between symptom onset and diagnosis, and total sleep duration, the cognitive gap between sleepy and non-sleepy patients persisted. Something else, the MAGNETO team concluded, must be driving both the lingering sleepiness and the cognitive toll that accompanies it.

Riese and Koehler believe they may have identified at least part of that missing something. Their hypothesis grew out of a detailed analysis of 543 patients with obstructive sleep apnea who had not yet begun any treatment. When the researchers split the cohort by employment status, a striking pattern emerged. Patients who held jobs scored significantly higher on the Epworth Sleepiness Scale, the most widely used questionnaire for measuring daytime sleep propensity, with a mean score of 11.4 compared with 9.7 in non-employed patients, a difference that reached statistical significance. Crucially, this was not because the employed patients had worse breathing disturbances. The apnea-hypopnea index, which counts the number of breathing interruptions per hour of sleep, was essentially identical between the two groups, at 37.8 versus 39.1 events per hour.

The real difference lay in how much time the employed patients actually had available for rest. Weekly rest time, the window between the end of work and the start of the next working day, was 68 minutes shorter in employed patients, averaging 443 minutes compared with 511 minutes in their non-employed counterparts. Accumulated across the week, this translated into a cumulative sleep debt of 82 minutes for working patients versus just 13 minutes for those not employed. In other words, two patients with the same severity of sleep apnea could arrive at the sleep clinic with very different sleep budgets simply because one of them had a job that consumed more of the hours available for recovery.

These findings echo data from a much larger European effort. The European Sleep Apnea Database, a multicenter registry of patients across the continent, previously reported that roughly 28 percent of patients with obstructive sleep apnea continue to experience excessive daytime sleepiness despite adequate continuous positive airway pressure therapy. That analysis attributed the residual sleepiness to a mixture of cultural and lifestyle factors rather than to treatment failure alone. The Marburg cohort now adds a concrete, quantifiable dimension to that interpretation: employment itself, through the mechanism of reduced rest time and accumulated sleep debt, appears to be one of those lifestyle factors with a measurable physiological footprint.

But the German authors argue that the pathway from employment to sleepiness involves more than simple arithmetic of lost hours. Drawing on the occupational health psychology literature, they point to a meta-analysis of 201 studies showing that people who experience a strong sense of calling in their work tend to work longer hours and, more importantly, achieve poorer psychological detachment, the mental ability to switch off from job demands during leisure time. A related study of high-commitment workers found that difficulty hanging up, both literally and figuratively, was linked to worse sleep quality and reduced morning vigor, independent of how many hours were actually slept. Psychological detachment, in other words, is a quality of recovery that a stopwatch cannot fully capture.

This distinction matters because the brain treats incomplete recovery as a form of sleep loss even when time in bed seems adequate. Decades of experimental work, including a widely cited meta-analysis by Lim and Dinges, have demonstrated that even short-term sleep restriction produces reliable impairments in attention, working memory, and processing speed, with vigilance tasks showing the most consistent and dramatic declines. If employed apnea patients are not only sleeping less but also detaching less effectively from work-related rumination, they may be compounding a subtle but cumulative cognitive burden. That burden would persist even after PAP therapy eliminates the breathing pauses, because the therapy addresses the airway but says nothing about the calendar or the smartphone buzzing with after-hours emails.

Connecting these threads produces a provocative reframing of the residual EDS problem. The MAGNETO study identified residual sleepiness as a cognitive risk phenotype, a subgroup of treated patients whose brains show signs of underperformance despite objectively adequate apnea control. Riese and Koehler suggest that this phenotype may partly reflect occupational sleep restriction rather than inadequately treated apnea. If that hypothesis holds up under prospective testing, it would carry immediate clinical consequences. Before escalating to wake-promoting agents, the stimulant medications sometimes prescribed for persistent sleepiness, clinicians might first take a careful occupational sleep history: how many hours does the patient work, does the job involve shifts or on-call duties, how much weekly rest time remains, and how well does the patient mentally disconnect from work in the evening?

The authors are careful to frame their data as hypothesis-generating rather than definitive. Their cohort was treatment-naïve, meaning the sleepiness measurements were taken before PAP therapy began, and the cross-sectional design cannot prove that employment causes residual sleepiness in treated patients. Employment status also correlates with many other variables, from body weight and comorbidities to socioeconomic factors, that could influence both sleepiness and cognition. What the letter does establish is a plausible, quantifiable, and, importantly, modifiable mechanism that has been almost entirely absent from the research agenda on residual EDS. Sleep debt of more than an hour per week, accumulated silently in the schedules of working patients, is exactly the kind of variable that large observational studies rarely capture.

To close that gap, Riese and Koehler propose a concrete research roadmap. Future prospective studies of residual sleepiness in PAP-adherent patients, they argue, should incorporate validated instruments measuring work hours, shift schedules, cumulative sleep debt, psychological detachment, and prosocial work orientation, the tendency to view one’s job as serving others, which intensifies the difficulty of switching off. Such measurements would allow researchers to disentangle how much of the heterogeneity in sleepiness phenotypes stems from occupational factors and how much from biological or treatment-related causes. They would also point clinicians toward a target that costs nothing to modify: restoring recovery time. For the growing population of treated sleep apnea patients who still struggle to stay awake, the most effective prescription may not come from the pharmacy but from a renegotiated boundary between work and rest.

Subject of Research: Occupational sleep restriction as a contributor to residual excessive daytime sleepiness in PAP-treated obstructive sleep apnea

Article Title: Occupational sleep restriction: a missing piece in residual EDS of PAP-adherent OSA?

Article References: Occupational sleep restriction: a missing piece in residual EDS of PAP-adherent OSA?. (n.d.). https://doi.org/10.1007/s44470-026-00153-0

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00153-0

Keywords: obstructive sleep apnea, excessive daytime sleepiness, PAP therapy, sleep debt, occupational health, psychological detachment, Epworth Sleepiness Scale, cognitive deficits, sleep medicine, work hours, CPAP, sleep restriction

Fragile X syndrome is caused by the loss or severe reduction of fragile X messenger ribonucleoprotein, commonly known as FMRP. This protein normally helps control the production of many other proteins at synapses, the microscopic junctions through which neurons communicate. When FMRP is absent, synaptic signaling becomes unusually responsive and poorly regulated. The resulting changes can affect learning, memory, attention, sensory processing and social behavior. In humans, fragile X syndrome is also frequently associated with anxiety, hyperactivity, autistic features and seizures, making the disorder biologically complex rather than a condition defined by a single cognitive impairment.

AM6545 belongs to a class of compounds that interact with the endocannabinoid system, a neuromodulatory network that helps regulate communication between neurons. The compound is designed to act on cannabinoid CB1 receptors, which are widely distributed throughout the brain and are involved in controlling the release of several neurotransmitters. Unlike cannabis-derived compounds that can directly activate CB1 receptors and produce psychoactive effects, AM6545 has been investigated as a negative allosteric modulator. In technical terms, it binds to a regulatory site on the receptor and changes how the receptor responds, potentially reducing excessive signaling while avoiding the full pharmacological profile of a conventional CB1 antagonist.

That distinction matters because the endocannabinoid system is deeply involved in synaptic plasticity, the capacity of neural connections to strengthen or weaken in response to experience. Synaptic plasticity is considered a cellular foundation for learning and memory. In fragile X syndrome, however, the balance between excitation and inhibition in neural circuits can be disrupted, and several studies have implicated abnormal endocannabinoid signaling in that imbalance. By modifying CB1 receptor activity, AM6545 may influence the release of neurotransmitters and help restore more stable communication within circuits involved in cognition. The new research indicates that this process is not limited to cannabinoid signaling itself.

The study by de los Reyes-Ramírez, Bergadà-Martínez, Martínez-Torres and colleagues identifies serotonergic signaling as a mechanism involved in AM6545’s pro-cognitive action in a mouse model of fragile X syndrome. Serotonin, or 5-hydroxytryptamine, is produced by a relatively small group of neurons concentrated mainly in the raphe nuclei of the brainstem. From there, serotonergic projections extend across broad regions, including the hippocampus and prefrontal cortex—areas essential for memory formation, attention and flexible decision-making. Serotonin acts through multiple receptor families, each capable of triggering distinct intracellular responses, so describing a “serotonergic mechanism” does not imply a single universal serotonin pathway.

Instead, the finding suggests that AM6545’s cognitive effects may depend partly on how the compound changes serotonin release, receptor activity or the responsiveness of neural circuits to serotonergic input. This is scientifically significant because serotonin and the endocannabinoid system are not isolated networks. They communicate at several levels: cannabinoid receptors can influence neurotransmitter release from serotonergic terminals, while serotonin receptors can alter the activity of circuits that also receive endocannabinoid regulation. Such cross-talk provides a plausible biological route by which a compound acting at CB1 receptors could produce effects that ultimately require serotonin signaling.

In animal research, evidence for this type of mechanism generally comes from combining behavioral tests with pharmacological or neurobiological interventions that selectively interfere with serotonergic transmission. When a compound improves performance in tasks measuring learning or memory, and that improvement is reduced or eliminated when serotonin receptors or related pathways are blocked, researchers can infer that serotonergic signaling contributes to the observed effect. The study’s central conclusion is therefore not simply that AM6545 changes behavior in a fragile X model, but that its cognitive benefit is linked to a specific neurotransmitter system. That distinction helps move the field from descriptive findings toward a more mechanistic understanding of drug action.

The implications extend beyond one experimental compound. Current clinical management of fragile X syndrome focuses largely on symptoms, using behavioral interventions and medications aimed at problems such as anxiety, attention difficulties, irritability, sleep disruption or seizures. No broadly approved treatment directly corrects the molecular cause of the syndrome or reliably restores cognitive function. A treatment strategy that engages both endocannabinoid and serotonergic biology could eventually offer new options, but the path from a mouse model to human therapy is long. Doses, metabolism, receptor distribution and safety margins can differ substantially between species, and improvements in laboratory tasks do not automatically predict meaningful gains in human learning or daily functioning.

The findings also raise questions that future studies will need to answer. Researchers will need to determine which serotonin receptor subtypes are essential, which brain regions are most important, and whether the response depends on the developmental stage or the severity of fragile X-related abnormalities. It will also be important to establish whether AM6545 improves multiple cognitive domains or primarily affects particular forms of memory, attention or behavioral flexibility. Long-term studies must examine tolerance, emotional effects, motor consequences and possible interactions with medications that already influence serotonin or cannabinoid signaling. These details will determine whether the mechanism is a promising therapeutic avenue or mainly a guide to understanding circuit dysfunction.

For now, the study offers a sharper view of how cognition may be altered in fragile X syndrome and how it might be pharmacologically improved. Its message is not that serotonin alone explains the disorder, nor that AM6545 is ready for clinical use. Rather, it reveals that the cognitive effects of manipulating cannabinoid receptor signaling can depend on communication with the serotonergic system. In a field where complex neurological symptoms arise from the interaction of many molecular pathways, that kind of mechanistic link is valuable. It gives researchers a new target for experiments—and a possible blueprint for designing treatments that aim not merely to suppress symptoms, but to stabilize the neural networks underlying learning and memory.

Subject of Research: Serotonergic mechanisms underlying the pro-cognitive effects of AM6545 in a mouse model of fragile X syndrome.

Article Title: A serotonergic mechanism is involved in the pro-cognitive effect of AM6545 treatment in a mouse model of fragile X syndrome.

Article References: de los Reyes-Ramírez, L., Bergadà-Martínez, A., Martínez-Torres, S. et al. “A serotonergic mechanism is involved in the pro-cognitive effect of AM6545 treatment in a mouse model of fragile X syndrome.” Translational Psychiatry (2026). https://doi.org/10.1038/s41398-026-04360-4

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41398-026-04360-4

Keywords: fragile X syndrome, AM6545, serotonin, serotonergic signaling, endocannabinoid system, CB1 receptor, cognition, synaptic plasticity, mouse model, neurobiology