Adherence to positive airway pressure in non-sleepy patients with obstructive sleep apnoea

To the Editor:

Complaints of excessive daytime sleepiness (EDS) are absent in many individuals with obstructive sleep apnoea (OSA). The influence of EDS prior to treatment on continuous positive airway pressure (CPAP) adherence has not been clearly determined [1, 2]. The aim of this prospective cohort study was to evaluate the adherence and perceived benefit during long-term CPAP therapy in a "real life" population of non-sleepy OSA patients.

Since May 15, 2007 consecutive patients investigated in seven sites in the west of France for symptoms of OSA (snoring, breathing pauses during sleep, morning headaches and EDS) have been recruited in the Institut de Recherche en Santé Respiratoire (IRSR) sleep cohort [2]. Patients filled in questionnaires including the Epworth sleepiness scale (ESS), the Pichot depression scale, the Medical Outcomes Study 36-item short-form (SF-36) for health-related quality of life (HRQoL), and questionnaires evaluating socioeconomic status (SES) [2]. As previously described [3], patients also completed surveys including anthropomorphic data and comorbidities. The study was approved by the University of Angers ethics committee and all patients gave their written informed consent. Between May 15, 2007 and May 15, 2012, we enrolled 4660 patients with an apnoea/hypopnoea index (AHI) ≥5 on sleep recording (overnight polysomnography or respiratory recording, performed as previously described [3]) of whom 1477 (31.7%) were non-sleepy (ESS <11 and no complaint of EDS) and 3183 (68.3%) were sleepy (ESS ≥11 or a report of frequently feeling unrested or sleepy). When compared with sleepy patients, non-sleepy patients were less likely to use anxiolytics, to be depressed or diabetic, had lower body mass index (BMI) and AHI. The percentage of patients with an education level ≤high school was higher in non-sleepy patients. On multivariable logistic regression analysis, including age, sex, BMI, AHI, depression, diabetes and level of education, only depression, AHI and level of education were independently associated with EDS. Adjusted odds ratios (95% confidence intervals) for being sleepy were 2.29 (1.97–2.68) for depression, 1.38 (1.21–1.58) for AHI ≥30 versus <30 and 0.81 (0.71–0.94) for a level of education ≤high school.

Among 2648 consecutive patients who were prescribed CPAP (fixed CPAP n = 980, or autoCPAP n = 1668; see [2] for details on CPAP initiation and follow-up) according to the French guidelines (AHI ≥30 regardless of symptoms or AHI ≥5 and <30 in combination with EDS that cannot be better explained by other factors or severe cardiovascular morbidity), 125 refused treatment, 135 were lost to follow-up and 2388 had available follow-up data, including 589 (24.7%) non-sleepy and 1799 (75.3%) sleepy patients of similar OSA severity (mean±sd AHI 40.8±19.0 and 41.0±22.5, respectively). The mean follow-up after initiating CPAP was 623±456 days. Data from questionnaires at diagnosis and during CPAP therapy at the most recent clinic visit are presented in table 1. In both sleepy and non-sleepy patients, CPAP therapy was associated with a significant improvement in ESS, in depression scale, and in four domains of HRQoL (energy/vitality, role emotional, role physical and social functioning). An improvement in three additional domains of HRQoL (general health, mental health and physical functioning) was observed in sleepy patients. Treatment satisfaction and mean CPAP side-effects score were similar in non-sleepy and sleepy patients. During follow-up, 466 (19.5%) patients discontinued CPAP after a mean duration of 319±294 days due to insufficient adherence (mean CPAP use 2.8±2.2 h·night⁻¹). On Kaplan–Meier analysis, only mild and non-significant differences were observed between non-sleepy and sleepy patients regarding the probability of continuing CPAP over time in the whole population, in patients with mild-to-moderate OSA (AHI <30, n = 603) and in patients with severe OSA (AHI ≥30, n = 1785). The average daily CPAP use recorded at each follow-up visit and the percentage of patients with mean daily CPAP use ≥4 h were also similar in non-sleepy and sleepy patients (table 1). On multivariable logistic regression analysis, including age, sex, BMI, AHI, depression, OSA symptoms, SES, CPAP mode and use of humidification, only AHI and the use of humidification were independently associated with CPAP dropout in non-sleepy patients. Adjusted odds ratios (95% confidence intervals) for CPAP dropout were 1.82 (1.13–2.92) for AHI <30 versus AHI ≥30 and 1.72 (1.29–2.31) for no humidification versus humidification.

In line with previous data [4], the present study demonstrates that EDS is more strongly associated with depression than with OSA severity. These results suggest that patients with clinical suspicion of OSA and EDS should be also assessed for depression. Sleepy patients also had a higher BMI and were more likely to be diabetic, but the independent association between EDS and metabolic disorders was not confirmed by the multivariable analysis. Education level may influence self-reported symptoms awareness and perception [5]. Although the direction of the relationship and the underlying mechanisms were not established in the present study, our findings suggest that a lower level of education is independently associated with a lower complaint of EDS in OSA.

It is often postulated that non-sleepy OSA patients are not good candidates for CPAP and will probably present poor adherence. However, previous studies evaluating the influence of EDS prior to treatment on CPAP adherence have reported conflicting results [1, 2]. The lack of interaction between OSA severity and sleepiness with regard to related cardiovascular morbidity suggests treating OSA patients with and without perceived EDS [6]. The vast majority of OSA patients without complaint of EDS were also found to exhibit attention defects that may potentially expose them to specific risks of accident during driving [7]. In line with recent findings from randomised controlled trials evaluating the impact of CPAP on cardiovascular outcomes in minimally symptomatic patients [8, 9], we demonstrate that long-term CPAP therapy is highly feasible and associated with an improvement in depression scores and HRQoL, in patients with an apparent paucity of symptoms at diagnosis. We also found that AHI is the only independent predictor of CPAP adherence prior to treatment in non-sleepy OSA patients. Altogether, our findings and those of recent randomised controlled trials [7, 8] indicate that the distinction between "OSA" and "OSA syndrome (OSAS)" may not be clinically relevant, since almost 460 patients that would not qualify for the definition of syndrome complied with CPAP and perceived treatment benefit. Our findings also suggest that the use of humidification may contribute to improving CPAP adherence.

One potential limitation of the present study is its observational design. However, rigorous observational cohort studies can complement randomised controlled trial data with information on long term treatment efficacy and adherence in patients encountered in day-to-day clinical practice [10]. The present study can be assumed to describe a “typical” pattern of OSA patients, as this multisite study included a large sample of patients with a wide range of disease severity. However, our study was carried out only in patients referred with clinical suspicion of OSA and some patients with AHI <30 were not prescribed CPAP according to the French guidelines. CPAP adherence and self-assessed benefit should, therefore, not be extrapolated to all subjects with undiagnosed OSA.

Despite these potential limitations, the present study demonstrates that CPAP therapy is highly acceptable and associated with self-reported benefits in non-sleepy OSA patients referred for sleep study. Our findings do not support the use of EDS measured by ESS and/or subjective assessment as a selection criterion for CPAP therapy in OSA patients.

• Received February 26, 2013.
• Accepted March 30, 2013.

• ©ERS 2013