Discussion
The purpose of the study was to assess SDB burden among I-DS referred to sleep center. All I-DS had OSA with majority exhibiting severe OSA. Hypoxemia and hypoventilation were also common. Among those who underwent AT, short term outcome shows marked improvement of OSA.
A high prevalence of OSA in children with DS has been described. A recent meta-analysis of 18 studies including 2,000 children with DS showed the prevalence of OSA to be about 50% by the conventional diagnostic threshold,(Lee et al. 2018) which represents a 10-fold higher risk than in the general pediatric population.(Marcus et al. 2012) Studies exclusively examining SDB in infant have been rare. Similar to our study, a previous single center study revealed virtually all I-DS who underwent sleep study manifested OSA and, alike our study, with a great proportion exhibiting severe OSA.(Goffinski et al. 2015) In our study severe OSA was present in 80% of I-DS. Thus, it is possible that the severity of OSA in infants may be higher than older children with DS.(Nerfeldt and Sundelin 2020) While the true prevalence of OSA is difficult to assess from clinical setting due to selection bias, it is reasonable to believe that the OSA risk is at its extreme even at a very young age in I-DS. Because all included infants met OSA criteria with most of them exhibiting severe OSA, we were unable to explore factors associated with OSA or severity of OSA. Not surprisingly, OSA was more severe in I-DS who underwent split study (vs. diagnostic) as young children with more severe OSA or anticipated to be more severe OSA tend to undergo O2 titration study in clinical practice.
In addition to OSA, CSA was also highly prevalent in I-DS. CSA during sleep is common in the preterm, newborn period and during infancy.(S et al. 2014) In healthy children, short duration (<20 s) CSAs in sleep are considered physiologic in the setting of a sigh, REM sleep and movement.(S et al. 2014) However, there is paucity of studies examining CSA in unselected infant population. Regardless, we suspect that high occurrence of CSA events noted in our study may not be unique to I-DS. A study by Fan et al. showed that CSA (as measured by CAI) was associated with younger age in the very youngest cohort (0-3 yrs old) in their review of children with DS who underwent PSG.(Fan et al. 2017) They reasoned that the improvement of CSA may be due to maturity of the respiratory control system with aging in younger children. However, such a relationship with age was not demonstrated in our study.
Sleep hypoxemia was also common affecting nearly 30% of I-DS in our study. As expected, hypoxemia was more common among those who exhibited hypoventilation and severe OSA in our study. Despite the notion that individuals with DS are at risk for hypoventilation, studies examining this very aspect in DS has been rare. A recent case control study by Richard et al. including children with and without DS revealed that children with DS have increased TCCO2 regardless of the presence of OSA and its severity.(Richard et al. 2020) In that study, investigators noted correlation between BMI and maximum CO2 speculating that obesity might be a contributing factor to the hypoventilation. Our study is the first to closely inspect hypoventilation exclusively in I-DS. In our study, 25% of I-DS met the criteria for hypoventilation. Interestingly, we found that I-DS with comorbid hypothyroidism had about 5 times higher odds of having hypoventilation, albeit borderline statistical significance likely related to small sample size. This alludes to the fact that hypotonia resulting from hypothyroidism, may play an important role in hypoventilation. The incidence of hypothyroidism in children with DS is 5.5%–10% but higher in the first year of life.(King et al. 2014) We did not examine the association of body habitus with hypoventilation given the limited variability of the BMI in our cohort.
In the subcohort of I-DS who underwent AT, we observed marked improvement in the severity of OSA as measured by oAHI and minO2sat. This is consistent with previous studies mainly including older children (older than infant age) that showed improvement of OSA following the AT.(Abdel-Aziz et al. 2017; Maris et al. 2017; Nerfeldt and Sundelin 2020) Average reduction in AHI was reportedly about 50% according to the systematic review.(Nation and Brigger 2017) Despite this, AT is not entirely curative in most cases and residual OSA is common.(Galluzzi and Garavello 2021) Moreover, complication rates of AT is higher in children with DS.(Goldstein et al. 1998; Yumusakhuylu et al. 2016) Thus, despite the marked improvement in OSA shown in our study, long term outcome and safety of AT in infants should be investigated in the future study.
OSA leads to poor sleep by frequent sleep disruption and deprivation of restful sleep. Poor sleep resulting from OSA in turn leads to daytime sleepiness, fatigue, mood change(Sánchez et al. 2009), and deficits in memory(Wallace and Bucks 2013), cognition(Pierobon et al. 2008), and executive function(Saunamäki et al. 2009), all of which may have more significant implications in I-DS given the inherent intellectual, cognitive, and emotional challenges associated with DS.(Sánchez et al. 2009) Moreover, gas exchange abnormality associated with SDB can unavoidably impact on the risk of cardiovascular disease including pulmonary hypertension in individuals with DS.(Bush et al. 2018) Thus, in view of our study findings, early screening of SDB as early as infants should be considered. This will become more feasible with the advance and emergency of more convenient technology that enables SDB evaluation in the home setting.
The strengths of this study are the exclusive investigation of infants and the comprehensive evaluation of all spectrum of SDB including hypoventilation. Moreover, to our knowledge this is the first study that examined the change in SDB metrics following the AT in I-DS. The main limitation of the study is related to inherent weakness of the retrospective study including selection bias. Such bias could have impacted the severity of SDB and the treatment outcome following the AT.
In conclusion, we report the excessively high burden of SDB among I-DS referred to tertiary sleep center. OSA was present in the entire cohort and was mostly severe in nature. In addition to OSA and CSA, sleep hypoxemia and hypoventilation were all common. There was significant improvement in OSA following AT in a subcohort. These findings propel us to consider early screening of SDB and possibly AT in this population.
Acknowledgement: We thank pediatric sleep specialists in our sleep center who cared for patients included in this study.
Grant support: YK is supported by grants from NIH (R01 HL158765, R21 AG070576, R21 HL150502).
Financial Disclosure: none. Non-financial or conflicts of interest that could be relevant in this context: none