[Cluster analysis of the clinical presentations in patients with moderate-to-severe obstructive sleep apnea syndrome and its significance].

OBJECTIVE

To analyze subtypes of the clinical presentation of moderate-to-severe obstructive sleep apnea syndrome (OSAS) by cluster analysis and to explore its significance.

METHODS

A retrospective analysis was performed on the data of adult moderate-to-severe OSAS patients, which were diagnosed in Respiratory Department, Renmin Hospital of Wuhan University from October 2012 to October 2015. Fifteen kinds of clinical presentations were included in the cluster analysis, and analysis of characteristic of each cluster was then performed after category.

RESULTS

Patients were classified as three distinct clusters. They were identified as "daytime sleepiness group" , "night insomnia group" and "minimally symptomatic group" , consisting of 450 cases (44.64%), 351 cases (34.82%) and 207 cases (20.54%) of the entire cohort. Members in daytime sleepiness group had the highest probability of daytime sleepiness, tiredness, irritableness and the highest Epworth Sleeping Scale score[329 cases (73.11%), 280 cases (62.22%), 223 cases (49.56%) and (13.50±4.93) points], members in night insomnia group had the highest probability of complaining of insomnia, nocturia, night sweats, apnea and awakening[177 cases (50.43%), 157 cases (44.73%), 130 cases (37.04%), 296 cases (84.33%), 182 cases (51.85%)], the probability of having typical symptoms above was the lowest in minimally symptomatic group, but these patients were more likely to have related comorbidities: hypertension, coronary heart disease, arrhythmia, chronic obstructive pulmonary disease and gastroesophageal reflux disease[118 cases (57.00%), 41 cases (19.81%), 44 cases (21.26%), 34 cases (16.43%), 38 cases (18.36%)]; all above results were significantly higher than the other two groups (P<0.012 5). The rate of diabetes in minimally symptomatic group[28 cases (13.53%)]was also higher but the difference was not statistical significance.

CONCLUSION

Cluster analysis contributes to classify multiple clinical presentations of OSAS patients, which has important significance for recognizing disease heterogeneity and early diagnosis.