Discussion
As with adults, there are all sorts of reasons why children do not sleep well. Some of those reasons are more serious than others. In our study, the sleep problem of the majority of the children (51.3%) was OSAS; however, only 7.4% of the total pediatric patients in the study underwent surgery for OSAS, even though some of the children with OSAS underwent surgery without undergoing preoperative PSG.
The etiologies of pediatric OSAS are multiple, and can be classified into intrinsic upper airway narrowing or increased upper airway collapsibility 4. Adenotonsillar hypertrophy is currently the most common cause of intrinsic upper airway narrowing, with other anatomical features resulting in upper airway narrowing including craniofacial syndrome, achondroplasia, Down syndrome, Beckwith Wiedemann syndrome, and MPS. In our study, three patients had underlying diseases with intrinsic upper airway narrowing: one had Prader-Willi syndrome, one had Down syndrome, and one had MPS. Adenotonsillectomy was suggested for these three patients, but their families refused this surgical intervention.
A decrease in muscle tone in the upper airway can cause increased upper airway collapsibility, with potential cause of such decreases including cerebral palsy, neuromuscular disorders, or inflammatory conditions such as allergic rhinitis and asthma. In our study, 2 OSAS patients had the underlying disease of Duchenne muscular dystrophy, and 24.5% (37/159) of the OSAS patients had allergic rhinitis or asthma.
Obesity appears to facilitate the emergency of OSAS; therefore, there is a high prevalence of OSAS among obese children. However, there is a higher proportion of children with OSAS who are obese. Thusm it appears that both OSAS and obesity can coexist and potentiate the adverse impacts of one another 5. The prevalence of OSAS is increasing globally due to the growing occurrence of obesity in society. In obese children, the fat deposits in the upper respiratory tract cause breathing difficulties during sleep, thus causing OSAS6,7, with reports of the prevalence of OSAS in obese children ranging from 13 to 59% 8. In our study, 3.1% (5/159) of the children with OSAS were also obese, and all of them were male patients. One of these 3 patients had severe OSAS and underwent an adenotonsillectomy, while it was recommended that the other 2 patients achieve body weight reductions through an increase in their dialy intense physical activities. A recent cross-sectional, prospective multicenter study, the NANOS study, assessed the contribution of obesity and adenotonsillar hypertrophy to pediatric OSAS and found that 46.6% of obese children in the community had OSAS 9.
Adenotonsillectomy is generally considered the first-line therapy in children with moderate or severe OSAS. In our study, there were 65 children with moderate OSAS and 9 with severe OSAS, and it was suggested that 26 of the moderate cases and all 9 of the severe cases undergo an operation. Of those 37 patients, 62% (23/37) underwent the operation. Some of the patients did not receive the operation because they their family members wanted to further observe their symptoms for a period of time. For these fourteen patients who did not undergo the operation, repeated PSG studies were performed for 3 of them who come back for follow-up visits, and their AHI results were found to be decreased in those follow-ups.
PSG prior to adenotonsillectomy is indicated for children with some conditions that increase the risk of perioperative respiratory complications. These conditions include obesity (especially if severe), Down syndrome, craniofacial abnormalities, neuromuscular disorders, sickle cell disease, or MPS 10. The purpose of the PSG in these high-risk children is to improve diagnostic accuracy and define the severity of OSAS to optimize perioperative planning. In our study, the OSAS children for whom the operation was suggested did have at least one of the aforementioned conditions, but some, including 3 with obesity, 2 with Duchenne muscular dystrophy, 1 with Down syndrome, 1 with MPS, and 1 with Prader-Willi syndrome did not undergo the operation.
Several authors have recommended that a clinical reevaluation be given to all children several months after adenotonsillectomy to determine whether snoring and the symptoms of OSAS have been resolved, especially in those children with higher risk of persistent disease, such as those with severe obesity or craniofacial syndromes. Furthermore, a postoperative PSG should be considered even in the absence of snoring or other symptoms in order to determine whether additional treatment is necessary for residual OSAS4. However, no studies to date have evaluated the timing of postoperative PSG evaluations, and this issue is not specifically addressed in the practice guidelines regarding the management of pediatric OSA. In our study, we did not collect postoperative PSG data, so further studies should be designed for evaluating this issue.
Medical therapies such as anti-inflammatory agents or CPAP are used as alternatives to adenotonsillectomy for children with OSAS, depending on the severity and specific locations of airway obstruction in the individual patient, and on associated comorbidities.
The second most common sleep problem among the children in our study was primary snoring, and the management of pediatric primary snoring consists of treating any upper airway obstructions or observation in cases in which there is no upper airway obstruction.
The third most common sleep problem in our study was PLMD. Thirty-one children (10%) had PLMD proven by PSG, and 7 of those 31 also had OSAS. In the large clinical case series reported by Gingras JL et al., PLMD was found to be common, affecting 14% of the 468 referred children11. It is relevant to note that there are no Food and Drug Administration-approved treatments for PLMD in children. However, most children with PLMD have low iron storage; therefore, iron therapy should be considered as the first line of treatment in children with PLMD whose iron levels are low. In our study, the iron level was checked in 9 children, and only one child with PLMD had a low iron level and was thus treated with iron therapy. Therefore, our management for PLMD consisted non-pharmacologic treatments, such as education, massage, exercise, or observation only.
The fourth most common sleep problem in our study was idiopathic hypersomnia, which is extremely rare in children. Hypersomnia is present in 4% to 6% of the general population, with only 1% of the population having idiopathic hypersomnia and most of the people with that being adolescents or adults 12. According to a report by Han F et al., 86% (361/417) of the children presenting with a complaint of primary hypersomnia to a sleep clinic in China met the criteria for narco­lepsy with cataplexy 13, while only 20% (3/15) of the children with excessive daytime sleepiness in our study had narcolepsy.
The clinical characteristics and experiences of CSA are very limited in children compared to the adult population, and it is thought to occur in about 1-5% of healthy children 14. CSA has been noted to occur more commonly in children with underlying diseases, and the presence of CSA may influence the course of those diseases14. In our study, there were 8 children who had CSA; 6 of them had secondary CSA and 2 had idiopathic CSA. Idiopathic CSA is really rare in children, and it cannot be reliably identified or diagnosed on the basis of history or a specific set of signs and symptoms 14. For the 2 children with a PSG-based diagnosis of CSA in this study, we suggested a magnetic resonance imaging (MRI) evaluation to assess for neuroanatomical abnormalities, but neither of the children came back for a follow-up visit.
A study by Felix O et al. reported that 18 of the 441 (4.1%) patients recorded during the study period had CSA, while 8 of the 310 (1.9%) patients in our study had central apnea 15. In the study by Felix O et al, the underlying disorders were dominated by neurosurgical disorders; however, congenital heart diseases dominated in our study.
Many parasomnias in children can be recognized by history alone, but some require nocturnal PSG for appropriate diagnosis and management. In our study, there were 7 patients who had enuresis, 5 who had bruxism, and 1 who had sleep walking according to the PSG results, with these 13 patients accounting for 4.2% of all the patients who underwent PSG. We believe however that there are still so many children with parasomnias who do not go to pediatric OPDs for help.
There were several limitations in our study. The first was that our study was only a 2-year retrospective study, and some children with sleep problems, especially those suspected of having insomnias and parasomnias, may go to psychiatric OPDs for help, not pediatric OPDs, and thus may not undergo a PSG study. Second, long-term follow-up is necessary for children with sleep problems in order to observe whether their symptoms are relieved or their AHI results are decreased after treatment, especially for those with OSAS after adenotonsillectomy. Third, blood sample tests, such as tests of serum Fe or ferritin levels, are not typically conducted for children with PLMD. Further studies prospectively collecting PSG data from children with pediatric sleep disorders will thus be required.