Women or men, between 20 and 65 years of age, presenting with a body mass index (BMI) of > 32 kg/m2 and daytime hypoventilation (ie, Paco2, > 45 mm Hg) in the absence of other known causes of chronic hypoventilation (eg, COPD [FEV1/vital capacity ratio, < 65%] or hypothyroidism) were eligible for the study. The study was approved by the hospital Ethics Committee, and patients gave written informed consent.
A diagnosis of OHS was established according to the diurnal Paco2 and pulmonary function test results. At baseline, patients also underwent overnight PSG testing. On the following morning, OSLER test and central CO2 chemosensitivity test were performed. Afterward, patients were referred to the pulmonary ward for 5 to 7 days in order to initiate therapy with NIV and to make adjustments to it. The same measurements were then performed with PSG recorded under NIV conditions.
PSG: At baseline, PSG was performed during spontaneous breathing in order to characterize the abnormal respiratory events associated with OHS. Sleep and respiratory events were recorded and scored manually according to standard criteria, as previously described.
REM hypoventilation was scored when progressive oxygen desaturation occurred that was associated with a sustained reduction in both flow and thoracic components of ventilation. During the same period, a constant or reduced respiratory drive (assessed by a reduction in respiratory effort as demonstrated by pulse transit time) should be observed without characteristic apneic or hypopneic episodes (Fig 1). Pulse transit time is a validated measure of respiratory effort. It has been validated against esophageal pressure. Thus, by semiquantitatively measuring respiratory effort, pulse transit time is a valuable measurement of the changes in respiratory drive occurring during REM sleep. In the definition proposed by Olson and Zwillich, hypercapnia is assumed to be present or to aggravate even the unmeasured transcutaneous Pco2.
Respiratory Function and Ventilatory Responses to CO2: Spirometry and plethysmography were measured according to the European Respiratory Society recommendations. CO2 chemo-sensitivity was assessed using Read’s method. The threshold of 1.5 L/min/mm Hg was used to separate low responders to CO2 from normal responders to CO2.
Subjective and Objective Sleepiness Assessment (Epworth Sleepiness Scale and OSLER Test): The epworth sleepiness scale is a validated eight-item, self-completion questionnaire. The OSLER test consisted of a 40-min sleep-resistance challenge that was conducted in a dark and quiet room. The subject was asked to remain awake and to react to a visual stimulus, which appeared for 1 s of every 3 s, by hitting a button. Sleep latency was the term used to describe the delay between the onset of the test and the moment corresponding to seven consecutive flashes (ie, 21 s) without response. Error profile 3-6 represented the number of three to six consecutive errors (ie, 9 to 18 s without a response from the patient), which is assumed to represent fluctuations in vigilance and microsleep episodes. Patients underwent an OSLER test at 9:00 am as we have previously described that this test has the most sensitive and specific criteria with which to detect sleepiness and impairment in attentional capabilities. Both in the study by Bennett et al and in our study, all of the control subjects were able to finish the 40-min test without falling asleep.
Patients were treated with bilevel positive-pressure ventilation (SERENA; SAIME; Savigny le Temple, France) in pressure support mode with a minimal respiratory rate setting. Inspiratory pressure was increased in order to achieve a maximal reduction in daytime Pco2 and optimal correction of nocturnal oxygen desaturation. Moreover, the control of nocturnal hypoventilation was assessed by measuring blood gas levels at the end of the night just before stopping NIV. Expiratory pressure was increased to eliminate obstructive sleep apnea.
A statistical software package (NCSS 97; NCSS; Kaysville, UT) was used for the statistical analysis. Results are expressed as the mean ± SD. A Wilcoxon test was used to compare measurements at baseline and when using NIV. We hypothesized that patients who had a low CO2 chemosensitivity would exhibit more REM sleep hypoventilation. OHS patients were then separated into groups of low and normal responders using a threshold of < 1.5 L/min/mm Hg. Variables between these two groups were compared using a Mann-Whitney test. The correlation between the proportion of time spent hypoventilating during REM sleep and CO2 sensitivity was assessed by a logarithmic best-fit analysis. For all tests, a significance level of 0.05 was used.
Figure 1. REM Sleep hypoventilation (a 5-min epoch is presented). SAT = arterial oxygen saturation; THE = buconasal thermistor; THO = thoracic movements; ABD = abdominal movements; FLO = nasal pressure; PTT = pulse transit time; EO1 = eye movements.