Abstract Body (Do not enter title and authors here): Background:
Expansion of thorax and abdomen during normal respiration generates negative intrathoracic pressure that affects organs like the heart and esophagus. Upper airway collapse during sleep increases respiratory effort. Measuring esophageal pressure (PES) is the gold standard for estimating this effort.
Heart failure is highly associated with obstructive sleep apnea (OSA). Abrupt swing in thoracic pressure and nocturnal hypoxemia during apnea are two mechanisms that can lead to heart failure.

Hypothesis:
PES and traditional sleep metrics: apnea hypopnea index (AHI), oxygen desaturation index (ODI), and mean-spO2 are associated with increased mortality risk in patients with heart failure and OSA.

Methods:
174 patients with heart failure identified in the local heart failure registry in Akershus university hospital. All patients underwent respiratory polygraphy and esophageal manometry during sleep. Observation time 2003-2024.
AHI, ODI, mean-spO2 and sleeping time with elevated esophageal pressure above different cmH2O, PES, were measured. Initial proBNP were also analyzed. Cox proportional hazard models were used to analyze the risk. Results were adjusted to age, sex, body mass index (BMI), and comorbidities.

Results:
Mean age 66 years (SD 12.019). 34 (20%) of observed cases were females. 70 (40%) patients died during observation time. Mean time to event (death) from first proBNP 3,5 years. Mean time to event from baseline polygraphy 8,2 years.
First proBNP predicted all-cause mortality (HR 1.108 [95% CI 1.052-1.167], p<0,001). AHI>15, ODI and mean-spO2 during sleep predicted all-cause mortality, but not sleep time with elevated esophageal pressure above 15 cmH2O. Sleep time with higher pressure degrees (above 80 cm H2O) did predict mortality.
Adding first proBNP to our prediction model did not affect the predictive value of mean-spO2 (HR 0.828 [95% CI 0.752 -0.912], p<0.001), or sleep time with PES above 80 cmH2O (HR 1.127 [95% CI 1.003-1.1.265], p 0.043).

Conclusion:
Nocturnal mean-spO2, and PES above 80 cmH2o predicted mortality in patients with heart failure and sleep apnea independent of first proBNP values.
Our results might explain why treatment with positive airway pressure is more effective in patients with severe OSA, and that oxygen therapy during sleep could represent a treatment choice for patients with HF and OSA.
Sleep studies may offer prognostic value for heart failure patients several years before first proBNP is obtained.