Identification of different clinical faces of obstructive sleep apnea in children

Yunxiao Wu; Guoshuang Feng; Zhifei Xu; Xiaodan Li; Li Zheng; Wentong Ge; Xin Ni

doi:10.1016/j.ijporl.2019.109621

Identification of different clinical faces of obstructive sleep apnea in children

Int J Pediatr Otorhinolaryngol. 2019 Dec:127:109621. doi: 10.1016/j.ijporl.2019.109621. Epub 2019 Aug 6.

Authors

Yunxiao Wu¹, Guoshuang Feng², Zhifei Xu³, Xiaodan Li⁴, Li Zheng⁴, Wentong Ge⁴, Xin Ni⁵

Affiliations

¹ Beijing Key Laboratory of Pediatric Otolaryngology, Head & Neck Surgery, Beijing Pediatric Research Institute, China.
² Research Center for Big data and Engineering, China.
³ Respiratory Department, China. Electronic address: zhifeixu@aliyun.com.
⁴ Otolaryngology, Head and Neck Surgery Department, Beijing Children's Hospital, Capital Medical University, National Center for Children's Health, Beijing, China.
⁵ Otolaryngology, Head and Neck Surgery Department, Beijing Children's Hospital, Capital Medical University, National Center for Children's Health, Beijing, China. Electronic address: nixin@bch.com.cn.

PMID: 31521054
DOI: 10.1016/j.ijporl.2019.109621

Abstract

Objective: This study aimed to identify the heterogeneity of obstructive sleep apnea syndrome clinical presentation in children.

Participants: Children who were 3-14 years old and with obstructive sleep apnea syndrome after polysomnography monitoring (apnea and hypopnea index>5 or obstructive apnea index>1) in the sleep center of Beijing Children's Hospital were included.

Methods: A sleep disorder questionnaire including different combinations of symptoms and co-morbidities of obstructive sleep apnea syndrome in children was used. A cluster analysis was used to classify the data.

Results: The apnea hypopnea index alone is not adequate to predict clinical phenotypes. Based on symptoms and co-morbidities of obstructive sleep apnea syndrome, three distinct clusters were identified. They were "nocturnal snoring and daytime sleepiness group" (cluster 1), "hyperactivity group" (cluster 2), and "minimally symptomatic group" (cluster 3). A prediction model was built according to eight variables which showed statistical significance by pairwise comparison among clusters. Overall accuracy of the prediction model could reach 86%. Both the sensitivity and specificity of cluster 2 and 3 prediction were around 90%.

Conclusion: Children with obstructive sleep apnea syndrome have different patterns of clinical presentation and the identification of the different clinical profiles of obstructive sleep apnea syndrome can provide clues for more personalised diagnoses and therapies.

Keywords: Children; Cluster analysis; OSA; Subtypes.

MeSH terms

Adolescent
Child
Child, Preschool
Comorbidity
Disorders of Excessive Somnolence / etiology*
Female
Humans
Male
Models, Statistical
Phenotype*
Polysomnography
Sensitivity and Specificity
Severity of Illness Index
Sleep Apnea, Obstructive / complications*
Sleep Apnea, Obstructive / diagnosis*
Sleepiness
Snoring / etiology*
Surveys and Questionnaires
Symptom Assessment