Search

Objective: To assess the effect of continuous positive airway pressure (CPAP) on blood pressure (BP), glucose and lipid profile, in women with moderate-to-severe obstructive sleep apnoea (OSA). Methods: We conducted a multicentre, open-label, randomized controlled trial in 307 consecutive women diagnosed with moderate-to-severe OSA (apnoea-hypopnoea index ≥15) in 19 Spanish Sleep Units. Women were randomized to receive effective CPAP (n=151) or conservative treatment (n=156) for 3 months. Systolic (SBP) and diastolic office BP (DBP), fasting glucose, glycated haemoglobin (HbA1c), insulin resistance measured by the homeostasis model assessment (HOMA) and total, LDL and HDL cholesterol were assessed at baseline and at the end of the follow-up. Data were analysed on an intention-to-treat basis, with adjustment for baseline values, body mass index, age, and specific medical treatment. The results are expressed as adjusted intergroup changes (95%CI). Results: Women had a mean (SD) age 57.1 (10.1) years, body mass index 34.3 (7.0) Kg/m2, ESS 9.8 (4.4) and apnoea-hypopnoea index 38.6 (20.8). Compared to the control group, the CPAP group achieved a significantly greater improvement in DBP (mean difference -2.04 mmHg; 95%CI -4.02 to -0.05; p=0.045), and a non-significant improvement in SBP (mean difference -1.54 mmHg, 95%CI -4.58 to 1.51; p=0.32). No differences were shown in fasting glucose, HbA1c, HOMA, and cholesterol measures between groups. Conclusion: In women with moderate-to-severe OSA, 3 months of CPAP therapy improves BP, but not the glucose and lipid profile, compared to conservative treatment.

Background. The development of home-based screening tests in sleep apnoea-hypopnoea syndrome (SAHS) diagnosis is essential to increase the accessibility to diagnostic facilities. Chronic obstructive pulmonary disease (COPD) is very prevalent among SAHS patients. However, almost all studies assessing home-based screening tests for SAHS excluded patients with COPD. Objectives. To assess the influence of suffering from COPD in the diagnostic performance of an at-home screening test for moderate-to-severe SAHS based on automated analysis of nocturnal oximetry by means of a multilayer perceptron neural network (MLP-NN). Methods. All patients showed moderate-to-high suspicion of suffering from SAHS. The population was divided into a training set (193) and two test sets (NO COPD: 110, COPD: 68). In-laboratory PSG was the gold standard. All patients carried out unsupervised nocturnal oximetry at home. Statistical, spectral and nonlinear features were computed and fed a regression MLP-NN trained to estimate the AHI. Results. The MLP-NN achieved similar concordance (intra-class correlation) between actual and estimated AHI in both NO COPD (0.73) and COPD (0.79) groups. The MLP-NN achieved high sensitivity values both in the NO COPD (97.5%, 80.9%) and in the COPD (86.5%, 84.6%) test sets for the cut-offs 15 and 30 events per hour, respectively. The MLP-NN reached lower specificity both in the NO COPD (24.1%, 70.2%) and in the COPD (37.5%, 69.0%) populations. Conclusions. Suffering from COPD does not affect the performance of a MLP-NN in the diagnosis of moderate-to-severe SAHS using unsupervised nocturnal oximetry at patient9s home.