Changes in behavioral and cognitive abilities after rapid maxillary expansion in children affected by persistent snoring after long-term adenotonsillectomy: A noncontrolled study

Introduction

The objective of this study was to verify changes in behavioral abilities and cognitive functions after rapid maxillary expansion (RME) in children with refractory sleep-disordered breathing (SDB) in the long term after adenotonsillectomy.

Methods

A prospective clinical trial study using RME therapy was conducted. Participant inclusion criteria were children who had adenotonsillectomy with maxillary transverse deficiency and persistent SDB (obstructive apnea-hypopnea index ≥1). The study included 24 children aged 5-12 years, and of these 24 children, 13 had primary snoring and 11 had obstructive sleep apnea. The patients underwent laryngeal nasofibroscopy and a complete polysomnography. In addition, patients completed the Obstructive Pediatric Sleep Questionnaire and Obstructive Sleep Apnea 18-Item Quality-of-Life Questionnaire. Behavioral and neurocognitive tests were also completed before and after RME.

Results

The Obstructive Pediatric Sleep Questionnaire and Obstructive Sleep Apnea 18-Item Quality-of-Life scores showed a statistically significant decrease in both groups ( P <0.001) after RME. The results showed that neurocognitive and behavioral parameters (Child Behavior Checklist scale) were similar in primary snoring and obstructive sleep apnea (OSA) groups before RME. In the OSA group, the mean scores of the “Somatic” and “Aggressiveness” domains decreased significantly ( P <0.05). The cognitive functions did not register significant differences pre- and post-RME in any of the cognitive functions, except for visuospatial function in the OSA group.

Conclusions

The noncontrolled design was a major limitation of our study. The need for treatment for SDB should consider the association of symptoms and behavioral disturbances with the child’s obstructive apnea-hypopnea index. RME might prove to be an alternative treatment for children with SDB refractory to adenotonsillectomy, improving quality of life and behavioral aspects. However, a larger sample size with a control group is needed to substantiate these claims.

Highlights

•

Orthodontics is a good approach for sleep-disordered breathing (SDB) in children.
•

Rapid maxillary expansion has a positive effect on pediatric SDB refractory to adenotonsillectomy.
•

Rapid maxillary expansion also improved the quality of life and behavioral aspects in children.
•

Symptoms and behavioral problems should be considered when choosing the therapy for pediatric SDB.

Sleep-disordered breathing (SDB) in children is common and described from the mildest form, namely primary snoring (PS), to obstructive sleep apnea (OSA), the most severe form based on polysomnography (PSG) indexes. Snoring is the most frequent complaint from parents seeking medical evaluation for children with SDB. Most children with OSA snore, although children who snore (8%-27%) do not necessarily have OSA. OSA is primarily characterized by persistent upper-airway obstruction rather than the total, intermittent obstruction seen in adults. Among children who snore, only 2% have OSA, and many do not have mouth-breathing patterns during the day, further complicating the diagnosis of SDB.

Children with PS and OSA may have behavioral, emotional, and neurocognitive difficulties. Depression, attention deficit, hyperactivity, aggressiveness, and impulsivity often lead to poor school performance. OSA may impair cognitive processing, especially sustained attention, memory, and learning disabilities in children. It is essential to emphasize the risk of impairment of behavioral and cognitive functions in children with milder forms of SDB. PS with daytime symptoms cannot be considered a benign condition.

The incidence of OSA in children with learning disabilities is 6-9 times higher than that of the general pediatric population. Several studies have shown a correlation between mouth breathing and difficulty in learning and disciplinary problems because of insufficient cerebral blood supply, causing sleepiness and consequently interfering with attention and possible impairment in understanding. On average, 67% of children diagnosed with OSA continue to have the symptoms as adults, which reveals the difficulty of early diagnosis and treatment in childhood and the possibility of repercussions in adulthood.

Adenotonsillectomy (AT) is considered the recommended first treatment line for children with OSA and can improve the apnea-hypopnea index, sleep quality, executive function of attention, and verbal ability. ^, Studies that compare those who undergo AT to a watchful waiting control group have shown that children who undergo AT exhibit a significant improvement in quality of life (QOL), which is a reliable predictor of behavioral outcomes. ^, The parent-reported symptoms using standardized symptom-based questionnaires of upper-airway obstruction were better indicators of most changes in children’s behavior than polysomnographic parameters. ^, OSA may recur (or persist) despite AT treatment, especially in children with underlying risk factors such as obesity and craniofacial disharmonies.

Rapid maxillary expansion (RME) is an orthopedic procedure that addresses OSA-related problems in children with maxillary transverse deficiency (MTD). Several studies have shown the short-term effectiveness of orthodontic treatment with RME, with evidence of significant improvement in OSA. RME may have a significant role in patients with OSA experiencing high, narrow palate associated with deepbite, retrusive bite, and crossbite. RME affects the nasopharyngeal airway by the transverse movement of the nasal lateral walls during expansion, significantly increasing the nasal cavity volume, reducing nasal air resistance, and facilitating physiological breathing. It can also positively influence the mandibular position, changing the size and volume of the oropharynx. RME can also improve snoring and the QOL of children with refractory SDB after AT.

There is an association between SDB with a negative impact on neurocognitive development. There has been growing evidence supporting the overlap between SDB and neurocognitive deficits in recent years. To our knowledge, few studies have assessed cognition and behavioral disorders in children with SDB. Therefore, this study aimed to explore the improvement effects of RME on behavioral abilities and cognitive functions in children with SDB refractory to long-term AT.

Material and methods

This prospective clinical trial of an orthodontic intervention was registered in the Brazilian Clinical Trials Registry: RBR-463by. Patients with residual snoring and MTD were recruited ≥2 years after AT. They underwent the following examinations before and after RME treatment: orthodontic documentation, laryngeal nasofibroscopy, and PSG. The Obstructive Pediatric Sleep Questionnaire (PSQ), ^, Obstructive Sleep Apnea 18-Item Quality-of-Life Questionnaire (OSA-18), ^, and neurocognitive and behavioral tests were applied in all children participating in the study.

Current evidence has shown that QOL is a confident prediction of behavioral outcomes. Data collected from 13 patients who participated in previously published studies ^, were used to calculate the sample size using data related to the primary outcome—the QOL of children before and after RME. Sample calculation was performed for each domain and the total score (OSA-18 questionnaire). Based on the results, the largest sample size obtained was chosen. Thus, adopting a study power of 80% and a type I error equal to 5% (α = 0.05), a sample size of 16 patients would be capable of identifying an average difference between pretreatment and posttreatment. The sample size was increased by 20% to allow for possible losses; thus, the sample size required for the study was at least 19 patients assessed before and after RME. The study design and participant flow are illustrated in Figure 1 .

The participants were recruited from patients attending the Pediatric Otorhinolaryngology Outpatient Clinic of Escola Paulista de Medicina at the Universidade Federal de São Paulo. The caretakers of 300 patients aged 5-12 years who had undergone AT in the previous 4 years at the University Hospital were contacted by telephone. Those who continued to present snoring symptoms at least 5 times per week were recruited. They were asked to attend a consultation with an otorhinolaryngologist who performed a laryngeal nasofibroscopy examination.

The laryngeal nasofibroscopy examination was performed with the child seated after the application of 2% lidocaine spray in the nasal cavities. Flexible fiber optics (Machida), a xenon light source (Styker, Othobean II), a film camera (Toshiba CCD IK M30AK), and a video monitor (Sony KV-CR) were used. During the examination, the presence of pharyngeal tonsils and the degree of obstruction caused by them were evaluated. Patients who required further surgery because of adenoid hypertrophy and lingual tonsillar tissue relapse were excluded. Patients with heart and neuromuscular diseases, craniofacial malformations, chromosomal syndromes, dental changes that prevent RME (loss of teeth, dental caries, or periodontal disease), and users of psychoactive drugs were also excluded.

The MTD diagnosis included the following methods: clinical evaluation, analysis of dental casts, and frontal cephalograms. Patients selected for treatment had a narrow maxilla (>5 mm MTD), high palate, unilateral or bilateral posterior crossbite, maxillary arch length discrepancy, decreased maxillary intermolar width, and transverse skeletal discrepancy between maxilla and mandible.

The Institutional Review Board approved the study under no. 0698011806/2017. All parents gave written informed consent after receiving information about the protocol, monitoring, and treatment proposed.

Twenty-four patients who met the study requirements were divided into 2 groups on the basis of pre-RME obstructive apnea-hypopnea index (OAHI) values, considering the following criteria: (1) PS group, OAHI <1 (n =13) and (2) OSA group, OAHI ≥1 (n = 11). Thus, the analyses were performed on the total sample and separately on the 2 groups.

After maxillary and mandibular alginate impressions, a hyrax expander was constructed for all patients with 2 bands, palatal stainless-steel bars with 1.0-mm diameter, and a jackscrew (Dental Morelli, São Paulo, Brazil) with stainless-steel extensions soldered to the palatal surfaces of each band. Each quarter-turn activation of the jackscrew was equivalent to 0.25 mm. The screw was activated (6 turns) to start the expansion, and the parents were instructed to perform 1 turn of the hyrax screw twice a day (0.5 mm per day) for the first 7 days. Then, the orthodontist reevaluated the patients, who would either decide to stop or continue to activate the appliance. Eventual additional activations were checked weekly and continued until the desired expansion was achieved. The expansion was stopped when the maxillary molars’ palatal cusp touched the mandibular molars’ buccal cusp. The device then remained in place for 6 months to allow bone formation in the palatal suture. All clinical treatments were performed by an experienced orthodontic consultant blinded to this research. After finishing the study, they proceeded with their orthodontic treatments to attain ideal transverse measurements and other parameters for excellent occlusion.

All patients underwent standard in-laboratory full-night PSG using digital PSG (Embla N7000; Embla Systems, Inc, Broomfield, Colo). The criteria adopted for sleep-wake scoring met the accepted international standards. Arousals were defined according to American Sleep Disorders Association criteria, whereas scoring of respiratory events followed the American Academy of Sleep Medicine criteria. Residual OSA was defined as an OAHI >1 per hour in addition to percutaneous oxygen saturation <92%.

The PSQ is a validated questionnaire comprising 22 items for screening patients with a history of OSA. All positive responses were grouped as “yes” and all negative responses as “no.” Each “yes” answer received a score of 1. ^, The PSQ scale can have clinical predictive value, provides valuable information, and is divided into 3 domains: snoring, sleepiness, and behavior. This questionnaire is useful as a complement to PSG in the assessment of children with OSA.

To assess the QOL of children with apnea, we applied the OSA-18 designed and validated in Portuguese. OSA-18 consists of 18 questions divided into 5 domains: sleep disturbance, physical symptoms, emotional symptoms, daytime function, and caregiver concerns. Each item has a score of 7 points (1, never; 7, always). ^, A score of <60 indicates a mild impact on QOL, 60-80 a moderate impact, and above 80 a serious impact.

Neuropsychologic assessment was performed by trained neuropsychologists. Each child was submitted initially to an assessment of intellectual performance by the Wechsler Abbreviated Intelligence Scale, a brief intelligence assessment scale that covers subjects aged 6-89 years. The instrument consists of 4 subtests: vocabulary, block design, similarities, and matrix reasoning. It provides information about the total, execution, and verbal intelligence quotients. However, an estimated total intelligence quotient can be obtained using only 2 subtests: vocabulary and matrix reasoning. In addition, the Brazilian version of the Child Behavior Checklist (CBCL) for children aged 6-18 years screening scale was used to assess participants’ social competence and mental health problems. The child’s main caretaker completed the questionnaire, when necessary, guided by the neuropsychologist to understand the questions correctly. The checklist is composed of 113 questions regarding manifesting behavioral problems and their frequency (0, absent; 1, occurs sometimes; 2, occurs often). The t-scores indicate the behavioral profile of children in 8 single indexes, namely anxious/depressed, somatic complaints (physical symptoms caused by emotional difficulties, such as dizziness, headaches, nausea, stomach pains, etc), social problems, thought problems, attention problems, rule-breaking behavior, and aggressive behavior. In the CBCL, a t-score ≥70 indicates the presence of behavioral and emotional problems at clinical levels.

To assess attentional processes, all participants underwent the Conners Continuous Performance Test (CPT-II). ^, CPT-II is a computerized instrument that consists of a presentation on the computer screen of different letters exposed at varying time intervals. The children were asked to press the left mouse button whenever any letter appears, except X, when they should not respond. The CPT-II standard paradigm consists of 6 blocks, with each block divided into 3 subblocks. The targeted and nontargeted stimuli (letters) are randomly shown for 250 milliseconds, with the interstimulus interval varying within each block. Tables of statistics (t-scores) for performance by block and for each of the interstimulus intervals (1, 2, and 4 seconds) are provided in the computer-generated report. It was adapted and validated for use with the Brazilian population. The range 45-59 classifies performance as average. A score >70 indicates an atypically slow response, and <40 is atypically fast.

The last test applied was the Developmental Neuropsychological Assessment. The test was translated, adapted, and validated for use in Brazil by Vargens and Argollo et al. It comprises 27 subtests to assess 5 cognitive functions: attention/executive function, language, visual-spatial processing, sensory-motor function and memory, and learning.

Statistical analysis

The sample was described as means and standard deviations (continuous variables: age, weight, height, and body mass index [BMI]) and frequencies (gender). For the comparison between groups, Student t tests were used for continuous variables and Fisher’s exact test for gender. The differences between pre-RME and post-RME were assessed by paired Student t tests.

A mixed analysis of variance was used to test the differences within subjects (time: pre-RME vs post-RME), between subjects (group: PS vs OSA), and the interaction (group × time). The interaction term (group × time) allows us to assess whether the differences over time (pre-RME vs post-RME) differ significantly between groups (PS vs OSA). To consider the age and height of subjects, the analysis of the variance model was adjusted for age and for the growth change between pre-RME and post-RME (both variables were included in the model as covariates). The Pearson correlation coefficient was used to study the correlation between cognition measures and the OSA-18 scores. Statistical analysis was performed with the SPSS software (version 26; IBM, Armonk, NY). A significant level of 5% was considered for the statistical tests.

Results

The sample comprised 24 white Brazilian patients of mixed ethnic ancestry, primarily males (n = 16, 66.7%), aged 6.1-12.7 years before RME. The mean age was 10.0 ± 1.8 years. On average, participants weighed 36.0 kg and were 1.36 meters tall. As expected, age, weight, and height significantly increased between the pre-RME and post-RME periods ( P <0.05) in the overall sample and within each group ( Table I ). As for age-adjusted BMI (Z-score for age), the mean decreased from 0.87 ± 2.16 before RME to 0.37 ± 1.76 after RME; however, the differences were not statistically significant ( P = 0.098). No significant differences were found between groups PS and OSA regarding age, weight, height, BMI, or gender ( P >0.05) ( Table I ).

Table I

Anthropometric parameters in children at baseline and 6 months after treatment

Variables	Baseline	6 mo after treatment	P value (Student t test) (pre-post)
Age, y
Total
Min-max	6.1-12.7	6.8-14.6
Mean ± SD	10.0 ± 1.8	11.3 ± 1.9	<0.001
PS
Min-max	8.0-12.7	9.2-14.6
Mean (SD)	10.3 ± 1.7	11.6 ± 1.8	<0.001
OSA
Min-max	6.1-12.1	6.8-13.7
Mean ± SD	9.7 ± 1.8	11.0 ± 2.1	<0.001
Group P values ^†	0.362	0.431
Weight (kg)
Total
Min-max	19-61	21-70
Mean ± SD	36.0 ± 9.0	40.9 ± 10.5	<0.001
PS
Min-max	27-61	33-70
Mean ± SD	39.2 ± 8.6	44.1 ± 10.0	0.002
OSA
Min-max	19-50	21-57
Mean ± SD	32.2 ± 8.2	37.2 ± 10.2	<0.001
Group P values ^†	0.053	0.110
Height (meters)
Total
Min-max	0.97-1.55	1.10-1.60
Mean ± SD	1.36 ± 0.14	1.46 ± 0.12	<0.001
PS
Min-max	0.97-1.54	1.40-1.60
Mean ± SD	1.37 ± 0.16	1.50 ± 0.07	0.003
OSA
Min-max	1.01-1.55	1.10-1.60
Mean ± SD	1.34 ± 0.13	1.41 ± 0.15	0.037
Group P values ^†	0.543	0.061
BMI (Z-score for age)
Total
Min-max	−2.13 to 5.21	−3.09 to 3.28
Mean ± SD	0.87 ± 2.16	0.37 ± 1.76	0.098
PS
Min-max	−1.84 to 5.21	−3.09 to 3.28
Mean ± SD	1.42 ± 1.94	0.50 ± 1.53	0.190
OSA
Min-max	−2.13 to 4.32	−2.88 to 3.16
Mean ± SD	0.23 ± 2.31	0.21 ± 2.07	0.969
Group P values ^†	0.182	0.699
Sex, n (%)
Total
Female	8 (33.3)
Male	16 (66.7)
PS
Female	4 (30.8)
Male	9 (69.2)
OSA
Female	4 (36.4)
Male	7 (63.6)
Group P values ^‡	1.000

SD , standard deviation; Min , minimum; Max , maximum.

† P value of Student t test for group comparison.

‡ P value of Fisher exact test.

The results of polysomnography values before and after RME are presented in Table II . The interaction (group × time) values were nonsignificant ( P >0.05) for all polysomnography variables, indicating no significant differences between groups regarding the differences between pre-RME and post-RME evolution.

Table II

Descriptive statistics and statistical comparisons of PSG data before and after RME

Group	Pre-RME	Post-RME	Mean difference	P value
Group	Pre-RME	Post-RME	Mean difference	Pre-post	Interaction (group × time)
AHI
Total	1.49 ± 1.32	2.07 ± 2.85	0.58	0.428
PS	0.83 ± 1.08	1.87 ± 3.65	1.04	0.392	0.401
OSA	2.26 ± 1.19	2.33 ± 1.41	0.07	0.834
Group P value	0.011 ^∗	0.567
OAHI
Total	1.07 ± 1.09	1.23 ± 1.47	0.16	0.845
PS	0.37 ± 0.30	1.08 ± 1.67	0.71	0.694	0.084
OSA	1.89 ± 1.12	1.42 ± 1.22	−0.47	0.364
Group P value	<0.001 ^∗	0.469
Basal SpO ₂%
Total	96.43 ± 0.94	96.29 ± 1.37	−0.14	0.864
PS	96.71 ± 0.87	96.82 ± 0.89	0.11	0.362	0.198
OSA	96.11 ± 0.96	95.60 ± 1.62	−0.51	0.958
Group P value	0.211	0.021 ^∗
Mean SpO ₂%
Total	95.60 ± 1.26	95.12 ± 1.67	−0.48	0.969
PS	96.00 ± 1.18	95.57 ± 1.66	−0.43	0.542	0.923
OSA	95.12 ± 1.22	94.54 ± 1.57	−0.58	0.714
Group P value	0.160	0.084
Minimum SpO ₂%
Total	90.83 ± 3.03	91.00 ± 2.75	0.17	0.165
PS	92.54 ± 1.81	92.15 ± 1.82	−0.39	0.616	0.562
OSA	88.82 ± 2.99	89.50 ± 3.10	0.68	0.418
Group P value	0.003 ^∗	0.035 ^∗