Obstructive Sleep Apnea-Hypopnea Syndrome Clinical in Subtypes A Principal Component Analysis-Based Cluster Analysis

Jiulong Kuang

DOI: https://doi.org/10.1016/j.chest.2016.02.591

IF: 9.6

2016-01-01

Chest

Abstract:SESSION TITLE: Sleep Disorders I SESSION TYPE: Original Investigation Poster PRESENTED ON: Saturday, April 16, 2016 at 11:45 AM - 12:45 PM PURPOSE: The clinical presentations of obstructive sleep apnea hypopnea syndrome (OSAHS) appear to be heterogeneous. OSAHS severity is usually classified using the apnea hypopnea index (AHI), which may not reflect the heterogeneity of distinct clinical manifestations of OSAHS. In this study, we investigate the clinical subtypes of OSAHS using multiple clinical variables. METHODS: A total of 308 patients examined by a full-night polysomnography were retrospectively analyzed. First, PCA was performed using ten variables selected for their relevance to OSAHS: age, body mass index (BMI), AHI, lowest oxygen saturation (LSaO2%), the longest duration time of sleep apnea (DBLon), arousal index, Epworth sleepiness scale (ESS), comorbidities, cumulative smoking and symptoms. Then, cluster analysis techniques based on PCA was used to identify subtypes of OSAHS. RESULTS: 255 men and 53 women with OSAHS aged 20～85 years were anal-ysed. Five main components were identified by PCA, which accounted for 73.091% of variance (table 1 and table 2). Four distinct clinical subtypes were identified using cluster analysis on the basis of PCA (table 3 and figure 1). Subtype1: relatively old patients with moderate OSAHS, shortest duration time of sleep apnea, had higher prevalence of comorbidities; Subtype 2: severe OSAHS patients, higher BMI, SaO2% was the lowest, had a higher arousal index; Subtype 3: you-nger patients with severe OSAHS, higher BMI, lower arousal index, lowest comorbidities, experienced more symptoms; Subtype 4: old patients, moderate OSAHS, the duration time of sleep apnea was the longest, high smoking index, minimally symptomatic patients. Importantly, patients with the same degree of AHI belonged to different subtypes, moreover, patients could manifest the similar clinical features and distributed in the same subtype, though their AHI was completely different. CONCLUSIONS: This study confirms the heterogeneity of OSAHS and combines AHI and other characteristics can better reflect the heterogeneity of OSAHS. CLINICAL IMPLICATIONS: This study confirms the heterogeneity of OSAHS and combines AHI and other characteristics can better reflect the heterogeneity of OSAHS. It is very important to identify the clinical subtype of patients with OSAHS, who may meet more personalized therapies. DISCLOSURE: The following authors have nothing to disclose: Jiulong Kuang No Product/Research Disclosure Information