Resting Energy Expenditure in Patients With OHS Treated With Long-term NIV

This study aims to compare, in subjects with obesity-hypoventilation syndrome (OHS) treated by long-term non-invasive ventilation (NIV), resting energy expenditure (REE) in spontaneous breathing and under NIV. The hypothesise of this study is that REE will be lower under NIV than under spontaneous breathing.

This is an observational monocentric study in which each subject will be its own control. Patients' resting energy expenditure will be measured at rest ("spontaneous breathing" condition) and under its usual treatment ("under NIV" condition) in a cross-over fashion. Conditions measurement order will be randomly generated by a computer algorithm such that half of the patients will be measured under spontaneous breathing before under NIV and the other half of the patients will be measured under NIV before under spontaneous breathing. Resting energy expenditure will be monitored during a short-term elective hospitalization (24 hours). Albeit for measurements of REE, tests which will be performed are part of the regular follow-up of patients under long term NIV. Primary objective is to compare in OHS patients the REE under spontaneous breathing and under NIV, using their usual settings. Description of the intended procedures and stages: 1. Recruitment and screening of patients fulfilling inclusion criteria will be done among patients with OHS under long-term NIV followed by our center (duration: 14 days). 2. Once informed written consent has been provided, patients will be included in the study protocol (time provided for decision: 28 days). 3. Patients are admitted for one night with overnight recording of nocturnal oximetry and transcutaneous capnography using their noninvasive ventilation at usual settings, prior to measurements of resting energy expenditure. Downloading of memorized data by ventilator software will also take place and will cover the previous month and the night of the study (duration: 12 hours). These procedures are part of the routine evaluation of patients under long term NIV. 4. Body weight will be measured, as well as body composition by bioelectrical impedance: briefly, a very low intensity electric current is sent between 2 electrodes placed on the dorsum of the foot and the ipsilateral hand. Measurement of resistance and reactance allows estimation of body compartments (fat-free mass and fat mass) (duration: 30 minutes). 5. Patients will undergo their first indirect calorimetric measurement. Prior to REE measurements, patients must be fasting for at least 10 hours, in a supine position, or if required for patient comfort, with a 30° inclination of head rest for at least 30 minutes. REE measurements are either performed with or without noninvasive ventilation (conditions order will be randomized by a computer)(duration: 30 minutes of rest and 20-30 minutes of indirect calorimetry). 6. Interval between measurements of at least 1 hour, to avoid impact of noninvasive ventilation on measurements. Between tests, they will perform pulmonary function tests and have their arterialized capillary blood gases sampled at earlobe (duration: 1 hour). These tests are part of the routine evaluation of patients under long term NIV. 7. Patients will undergo their second indirect calorimetric measurement (see details in section on indirect calorimetric). Prior to REE measurements, patients must be fasting for at least 10 hours, in a supine position, or if required for patient comfort, with a 30° inclination of head rest for at least 30 minutes. REE measurements are either done with or without noninvasive ventilation (randomized order by a software) (duration: 30 minutes of rest and 20-30 minutes of indirect calorimetry). 8. These two tests will be performed in a randomized order (spontaneous breathing or noninvasive ventilation). 9. Information to the patient of the results, which will also be forwarded to physician and nutritionist in charge. In 2016-2017, seventy-four OHS patients treated at home by long-term NIV were identified in our area. Expecting a participation rate of around 50%, at least 30 patients will be included in the present study. With such a sample size, the study will have a power of 80% to detect an effect size of 0.5.