基于6~8岁替牙期儿童颜面部特征及临床症状的重度腺样体肥大预测模型

A prediction model for severe adenoid hypertrophy based on craniofacial characteristics and clinical symptoms in 6-8 year old children during mixed dentition stage

摘要:

目的 针对替牙期儿童重度腺样体肥大(AH)与颅颌面发育异常的关联，研究无创面容测量、临床症状与AH的相关性，并构建重度AH早期诊断的预测模型。方法 采用横断面研究设计，纳入2023年8月—2024年12月于上海第九人民医院就诊的6~8岁替牙期儿童201例。收集儿童颅颌面形态参数：侧面照片颜面测量角度及比值；临床症状指标包括：呼吸模式、扁桃体肥大程度、反复扁桃体发炎史、鼻炎/哮喘患病情况、OSA-18问卷得分、中耳炎史等。按照3∶1比例将研究对象分为建模组149例，验证组52例。基于鼻内镜检查将建模组分为轻/中度AH组(n=77)与重度AH组(n=72)，采用LASSO回归筛选阳性变量，构建Logistic回归预测模型，并应用ROC曲线(AUC)、校准曲线及决策曲线分析(DCA)验证模型效能。结果 与轻/中度AH儿童相比，重度AH儿童呈现面凸度减小[(173.24±2.71)° vs (171.01±4.08)°，P<0.001]、唇突比值增加(0.90±0.09 vs 1.02±0.11，P<0.001)、上颌凸度增加[(30.37±6.52)° vs (35.98±7.25)°，P<0.001]等特征性面容改变。呼吸模式(OR<1，P<0.05)、慢性扁桃体炎(OR=6.035，P=0.007)、慢性鼻炎(OR=5.183，P=0.013)、哮喘(OR=14.927，P=0.002)、重度鼾症(OR=5.803，P=0.011)也与重度AH之间存在关联性。基于上述因素构建的复合预测模型在建模组AUC=0.949，验证组AUC=0.961，提示该模型具有良好的区分度和临床适用性。结论 本研究发现替牙期儿童的面凸度、唇突比值、上颌凸度等面容特征是重度AH的早期敏感指标。基于此构建的面容-症状预测模型提供了一种无创、高效的筛查方法，可以有效帮助重度AH的临床前诊断及广泛筛查。

Abstract:

Objective Regarding the association between severe adenoid hypertrophy (AH) and craniofacial development abnormalities in children during the mixed dentition period, to investigate the correlation between non-invasive facial measurements, clinical symptoms and AH, and construct a predictive model for the early diagnosis of severe AH.Methods A cross-sectional study was adopted, including 201 children aged from 6 to 8 years in the mixed dentition stage who visited Shanghai Ninth People’s Hospital from August 2023 to December 2024. The craniofacial morphological parameters including facial measurement angles and ratios from lateral photos, data of clinical symptom indicators including breathing patterns, degree of tonsillar hypertrophy, history of recurrent tonsillitis, history of rhinitis/asthma, OSA-18 questionnaire scores, history of otitis media, etc were collected. All the study subjects were randomly divided into a modeling group (n=149) and a validation group (n=52) at a ratio of 3∶1. The modeling group was further divided into a mild-to-moderate AH group (n=77) and a severe AH group (n=72) based on nasal endoscopy. LASSO regression was used to screen positive variables, and a Logistic regression prediction model was constructed. The model performance was verified via receiver operating characteristic (ROC) curves, calibration plots, and decision curve analyses (DCA).Results Compared to the mild-to-moderate AH group, children with severe AH presented with decreased facial convexity [(173.24±2.71)° vs (171.01±4.08)°,P<0.001], enhanced lip prominence (0.90±0.09 vs 1.02±0.11, P<0.001),and increased maxillary convexity [(30.37±6.52)° vs (35.98±7.25)°,P<0.001], as well as other characteristic facial changes. Multivariate analysis revealed that the breathing patterns (OR<1, P<0.05), chronic tonsillitis (OR=6.035, P=0.007), chronic rhinitis (OR=5.183, P=0.013), asthma (OR=14.927, P=0.002), and severe snoring (OR=5.803, P=0.011) were also associated with severe AH. The composite prediction model based on these factors had an AUC of 0.949 in the modeling group and 0.961 in the validation group, indicating good discrimination and clinical applicability.Conclusions The facial convexity, lip prominence, and maxillary convexity in children during the mixed dentition stage are early sensitive indicators of severe AH. The facial-symptom prediction model constructed based on these factors provides a non-invasive and efficient screening method that can effectively assist in the preclinical diagnosis and wide screening of severe AH.