Relationship between obstructive sleep apnea and attention deficit hyperactivity disorder in children
刘海琴,刘灵,袁钰淇,任晓勇,刘娜,张一彤,冯雅妮,折宁宁,施叶雯,思超
DOI: https://doi.org/10.3760/cma.j.cn115624-20210324-00150
2021-01-01
Abstract:Objective:To analyze the correlation between obstructive sleep apnea (OSA) and attention deficit hyperactivity disorder (ADHD).Methods:The clinical Data, polysomnography (PSG) and cognitive function examination results of 112 OSA children admitted to Department of Otorhinolaryngology Head and Neck Surgery of the Second Affiliated Hospital of Xi′an Jiaotong University from January 2019 to June 2020 were retrospectively analyzed. According to the severity of OSA, the children were divided into mild, moderate and severe OSA groups, and the basic demographic characteristics, sleep parameters and ADHD occurrence were analyzed. According to the results of ADHD examination, the children were divided into ADHD group and non-ADHD group, and the basic demographic characteristics and sleep parameters were analyzed. Taking these parameters as independent variables, binary Logistic regression analysis was conducted to establish the model equation for predicting the risk of OSA associated ADHD among children.Results:Grouped by OSA severity, among the three groups, apnea-hypopnea index (AHI) [3.70 (2.84, 5.47) vs 8.59 (7.50, 9.54) vs 19.48 (15.83, 25.23)], obstructive apnea index (OAI) [1.31 (0.93, 1.82) vs 3.03 (1.54, 4.41) vs 11.69 (8.53, 15.42)], obstructive apnea-hypopnea index (OAHI) [2.82 (1.81, 3.64) vs 6.17 (5.58, 7.26) vs 15.68 (13.12, 21.25)], and respiratory event-related arousal index [0.50 (0.25, 1.05) vs 1.25 (0.70, 2.23) vs 2.40 (1.60, 4.70)] increased, minimum pulse oxygen saturation (SpO 2) [90.00 (88.00, 92.00) vs 87.00 (83.00, 90.25) vs 81.00 (76.00, 85.00)] decreased, the differences were statistically significant (all P<0.05). The non-rapid eye movement (NREM)1 period time ratio of the severe OSA group was significantly longer than that of the mild OSA group, while the average SpO 2 was significantly lower than that of the mild OSA group; the NREM3 period time ratio of the moderate and severe OSA group was significantly less than that of the mild OSA group; the arousal index of the severe OSA group was significantly greater than the mild or moderate OSA group. There were no statistically significant differences among the three groups in gender, age, body mass index, sleep efficiency, rapid eye movement (REM) period time ratio, and NREM2 period time ratio (all P>0.05). Mild OSA group had 10 cases of ADHD (17.54%), moderate OSA group had 7 cases (23.33%) of ADHD, severe OSA group had 9 cases of ADHD (36.00%), and the difference was not statistically significant. Grouped by ADHD examination, the AHI, OAI, OAHI, and NREM1 period time ratios of the ADHD group were significantly higher than those of the non-ADHD group, while the sleep efficiency, minimum SpO 2 and NREM3 period time ratio were significantly lower than those of the non-ADHD group. The Logistic regression analysis suggested that ADHD was correlated with sleep efficiency, minimum SpO 2, and NREM3 period time.The established Logistic regression equation was: X=15.670+0.061×(sleep efficiency)-0.212×(minimum SpO 2)-0.144×(NREM3 period time ratio), the sensitivity and specificity of the model prediction were 84.6% and 79.1% respectively when the area under the receiveroperating characteristic curves was 0.867. Conclusions:OSA and ADHD in children have a certain correlation. Sleep structure disturbance and intermittent hypoxia may be important reasons. The predictive model equations obtained by PSG in this study can be used to assess the risk of ADHD in children with OSA.