Abstract Rationale Obstructive sleep apnea (OSA) is a prevalent disorder characterized by repetitive upper-airway obstruction and loss of ventilation due to reduced pharyngeal muscle activity. Pharmacologic approaches to maintain muscle activity are promising, with atomoxetine-plus-oxybutynin (“AtoOxy”, or its commercial analogue AD109) yielding encouraging findings in recent trials. AtoOxy is thought to activate pharyngeal muscles, via noradrenergic stimulation and antimuscarinic disinhibition at the hypoglossal motor pool. However, whether atomoxetine and oxybutynin work synergistically as stimulants and which one of these muscles they predominantly exert their effect on has yet to be demonstrated. Methods In a crossover mechanistic trial, 20 participants with suspected or diagnosed OSA underwent overnight polysomnography measuring ventilation (oronasal mask), intramuscular genioglossus (GG) EMG, intramuscular tensor palatini (TP) EMG, and ventilatory drive (diaphragm EMG catheter). Across three nights, participants received AtoOxy (80/5 mg) or atomoxetine alone (random order), followed by placebo. Interventions were administered ∼60 min after sleep onset to provide a within-night control period. Mixed-model analysis compared the effects of each intervention on ventilation, GG EMG and TP EMG (peak, %max) at minimum ventilatory drive (1st decile), when the airway is most vulnerable. Results Nineteen participants (10F:9M, age=44 34, 54 y, BMI = 30 26, 93.5 kg/m2) provided data for analysis (16 completed both drug-night visits, 12 completed the entire study). Compared to placebo, AtoOxy improved ventilation (+45.6 34.5, 56.6 %eupnea, estimate 95%CI), indicating enhanced airway function. AtoOxy did not increase genioglossus activity (peak −1.1 −2.6, 0.4 %max) but markedly increased TP activity (peak +3.9 2.1, 5.7 %max). Atomoxetine alone showed smaller effects on ventilation (−11.8 −21.8, −1.7 %eupnea, Pdifference= 0.021), but similar effects on muscle activity (Pdifference0.9). Sensitivity analysis at eupneic ventilation detected a reduction of GG with AtoOxy (−2.0 −3.1, −0.9 %max, P = 0.0005), while ventilation and TP increased. Despite this, GG responsiveness to increasing drive was augmented with AtoOxy (1.3 0.6, 2.1 %max/drive, P 0.0001) and atomoxetine alone. TP responsiveness was increased with AtoOxy but not atomoxetine alone (Pdifference0.0001). Conclusion AtoOxy improves pharyngeal function while raising TP activity but not increasing GG activity (mainly due to atomoxetine’s effect). Contrary to prevailing interpretation, the physiological mechanism of AtoOxy likely lies with the restoration of wakefulness-like input to non-GG muscles, rather than direct GG activation, which declines as airway patency improves. This mechanistic insight supports development of drugs directly targeting GG activation, identifying AtoOxy responders as those relying most on non-GG muscles, and combining AtoOxy with other GG-stimulatory interventions. This abstract is funded by: NIH
Monteiro et al. (2026) studied this question.