This IRB-approved, retrospective review examined children with T21 who were 60 months old or younger at the time of first PSG between 2015 and 2020 and included both symptomatic and asymptomatic patients. Eligible patients were identified via review of EMR (electronic medical record) with problem list containing ICD code for Trisomy 21 (Q90.0) at the time of the PSG and confirmed via manual chart review by a member of the research team.

Data collected for this study consisted of demographic information, relevant medical history, PSG data, and treatment information. All data were extracted retrospectively from the EMR via chart review by a member of the research team. PSG data was collected via sleep study report available in the EMR. All patients included in this study completed a clinical office visit with a pediatric sleep specialist prior to the PSG which includes clinical exam and documentation of relevant medical and surgical history. These visits are typically completed <2 months prior to the PSG. Clinical and demographic data were collected for this project via review of the problem list and documentation in the EMR at the time of the office visit.

The demographic information consisted of age, sex, race, ethnicity, height, height z-score, weight, weight z-score, BMI, BMI z-score, medications at the time of the PSG, and tobacco smoke exposure. Z-scores for height and weight were determined using Trisomy 21 growth charts while z-scores for BMI were determined using WHO growth charts.

Relevant medications included nasal steroids, antihistamines, montelukast, melatonin and other non-prescription sleep aids, inhaled medications, histamine-2 antagonists (hist-2), and proton pump inhibitors (PPI). These data were collected from the EMR as documented in the medication review at the clinical encounter immediately prior to the PSG. Relevant medical history included the existence of relevant pre-existing conditions, feeding tube prior to the first PSG, ENT visits, ENT surgeries prior to the first PSG, ENT surgery at any time, prematurity, pulmonary conditions, dysphagia and other gastrointestinal conditions, cardiac disease, thyroid conditions, and neurodevelopmental conditions. The comorbidities were defined based on review of the problem list in the EMR at the time of the visit as well as documented in the clinical notes at the time of the clinical visit immediately prior to the PSG. Prematurity was defined as being born prior to gestational age of 37 weeks. Pulmonary conditions consisted of (1) history of asthma or recurrent wheezing that was diagnosed by a physician or for which the patient was prescribed pulmonary inhaled medications, (2) bronchopulmonary dysplasia (BPD) or chronic lung disease of prematurity, and (3) structural airway abnormalities, specifically oropharyngeal malformations, laryngomalacia, laryngeal clefts, vocal cord paralysis, tracheomalacia, bronchomalacia, and tracheoesophageal fistula (TEF). Gastrointestinal conditions consisted of (1) reflux that was diagnosed by a physician or for which the patient was prescribed reflux medications, (2) gastrointestinal surgical repair, and (3) structural GI anomalies, specifically duodenal atresia, duodenal stenosis, TEF, intestinal malrotation, and Hirschsprung's disease. Cardiac conditions consisted of (1) pulmonary hypertension, (2) surgical repair of structural heart defects, and (3) congenital structural heart defects, specifically ASD, VSD, AV canal, TOF, coarctation of aorta, and PDA. PFOs were not included as congenital structural heart defects. Thyroid conditions consisted of hypothyroidism and hyperthyroidism. Neurodevelopmental conditions consisted of autism and ADHD. ENT surgery included myringotomy tube placement, adenoidectomy, tonsillectomy, adenotonsillectomy, and supraglottoplasty.

PSG was performed in an American Academy of Sleep Medicine-accredited sleep laboratory at Nemours Children's Hospital, Delaware. PSG is the gold standard for the diagnosis of OSA (17). Electroencephalography, electrooculography, electromyography, and electrocardiography were continuously recorded throughout the PSG. Respiratory effort was measured via respiratory inductance plethysmography and oxygen saturation was measured via a finger probe on a pulse oximeter. A nasal pressure cannula and a thermistor measured airflow, and snoring was measured via a microphone. The patients were monitored using by a polysomnographic technician during the entire duration of the study in a dark and comfortable environment. The raw data was reviewed and interpreted by a pediatric board-certified sleep medicine physician. Respiratory events were scored according to the American Academy of Sleep Medicine guidelines using pediatric scoring (19).

The PSG data collected included respiratory and sleep architecture data. The respiratory parameters consisted of the central apnea index (CAI), obstructive apnea index (OAI), hypopnea index (HI), apnea-hypopnea index (AHI), number of episodes of oxygen saturations less than 90% (# Episodes SpO₂ < 90%), the percent of total sleep time spent below 90% saturation (TST with SpO₂ < 90%), number of episodes of relative oxygen desaturations ≥3% (Episodes SpO₂ desaturation ≥ 3%), SpO₂ nadir, mean SpO₂ in REM, mean SpO₂ in NREM, peak End tidal CO₂ (EtCO₂) level, percent of total sleep time spent with a peak EtCO₂ level above 50 torr (TST w/EtCO₂> 50 torr), periodic breathing time (PBT), percent of total sleep time spent with periodic breathing (TST with PBT), mean respiratory rate in REM (MRR REM), and mean respiratory rate in NREM (MRR NREM), periodic limb movements of sleep index (PLMI), periodic limb movements associated with arousals (PLM assoc. with arousals) for children above the age 2 years. Sleep architecture included the patient's total sleep time (TST), sleep efficiency (SE), sleep latency, percentage of TST in REM, percentage of TST in N3, and arousal index. Sleep efficiency was considered decreased if it was <90%. Severe sleep apnea was noted if the AHI was ≥10/hour. Potential treatment options were supplemental oxygen, airway surgery, positive airway pressure (PAP), intranasal corticosteroids, and other. Airway surgery consisted of adenoidectomy alone, adenotonsillectomy, and supraglottoplasty.

All analysis was performed using group-generated R scripts. Descriptive statistics consisted of mean, median, standard deviation, and percentages where appropriate. In subgroup analysis, the younger (0–35 months) patients were compared to the older (36–60 months) patients using 2-sample Wilcoxon-rank sum test after tests for normality were performed using the Shapiro-Wilk test. A p value of ≤0.05 was considered statistically significant.

The full cohort was predominantly female (54.7%), white (61.9%), and non-Hispanic or Latino (71.9%). 14.3% of the cohort were black, and 22.2% were listed as “other.” The mean age of the participants was 30.8 months (median = 30.5, SD = 16.4). See Table 1 for descriptive statistics for height, weight, and BMI and z-scores for those parameters. 63.5% were taking relevant medications with the most common being inhaled respiratory therapies (39.7%). Roughly a quarter had a history of tobacco smoke exposure.

The most prevalent comorbidities include congenital heart disease (78.1%) and asthma/wheezing history (43.8%). Table 2 details the percentage of patients with each comorbid condition.

In this cohort, many children were referred for PSG with history of airway concerns including enlarged tonsils (62.5%), snoring (81.2%), and restless sleep (59.4%). 18.8% were considered to be asymptomatic and referred due to age and history of Trisomy 21. Noisy breathing was present in approximately 25% of the population. Asthma or wheezing was present in 43.8% of patients of which most (89.3%) were treated with inhaled medications. Reflux was diagnosed in 64.6%, and 42.9% were treated with medical therapy. Roughly half of the patients had a feeding tube (either nasogastric tube or surgically-placed tube) at some point prior to the PSG.

Table 3 contains respiratory parameters of the PSG. In general, these patients had high rates of obstructive apnea and hypopnea events (mean = 21.4, median = 17.8, SD = 19.1). The hypopnea index is high (HI: mean = 15.2, median = 12, SD = 16.4). The SpO₂ nadir was moderately reduced with a mean value of 86.2% (median = 88.5%, SD = 7.8%); however most had saturations above 90% for majority of the study (Duration of SpO2 below 90%; mean = 1.96% of TST, median = 0% of TST, SD = 7.33% of TST). Mean oxygen saturation recorded was 97% (median = 97%, SD = 2.1% in REM; 1.7% in NREM). In addition to experiencing frequent apnea and hypopnea episodes, the patients also demonstrated borderline abnormal ventilation an elevated mean peak EtCO2 (mean = 55.9 torr, median = 54.8 torr, SD = 6.75 torr) and percentage of time above 50 torr (mean = 20.8%, median = 5.6%, SD = 29.4%).

Table 3 includes all descriptive statistics for the cohort's sleep architecture. The majority (94%) had TST >300 min. 78.1% of studies showed decreased sleep efficiency of <90%. The sleep latency (mean = 21.4 min, median = 9.25 min, SD = 26.5 min) was within normal range, but the arousal index (mean = 32, median = 29.2, SD = 17.7) was elevated (20). The patients spent more time in N3 (mean = 42.7%, median = 41.1%, SD 15.1%) and less time in REM (mean = 14.4%, median = 13.6%, SD = 6.4%) than expected for the mean age (20).

A majority of the cohort was found to have OSA (84.1%), and 67.2% had severe sleep apnea (defined as AHI >10 events/hour). Airway surgery was the most common treatment choice (65.1%) followed by intranasal corticosteroids (11.1%), PAP therapy (6.4%), other (4.8%), and supplemental oxygen (3.2%). Many patients (12.5%) received more than one treatment modality. Approximately 30% of patients were followed with clinical observation after the baseline PSG.

The entire cohort was sub-divided into younger (0–35 months of age) and older (36–60 months of age) cohorts with 35 and 29 patients, respectively. There were no significant differences in the majority of PSG parameters and specifically no differences in rates of OSA. See Table 3 for full comparison of all PSG parameters between the two groups. One notable difference between the two group's PSG parameters was in the number of desaturations less than 90% with the younger cohort having a mean of 42.2 compared to the older group having a mean of 7.8 (p = 0.05). As expected physiologically, the mean respiratory rates were higher in younger cohort in both REM (21.3 vs. 15.6; p < 0.01) and NREM (20.8 vs. 17; p < 0.01) Another important difference between the two cohorts was the diversity of treatment modalities. The older cohort almost exclusively underwent an AT or clinical observation with only 3.45% of patients receiving an alternative strategy. In comparison, 28.6% of the younger cohort received treatment other than an AT or clinical observation (including supplemental oxygen, CPAP).