The authors of this study aimed to evaluate the use of polysomnography (PSG) in children with Down syndrome (DS) between ages 0 and 7 years, to assess the prevalence and severity of obstructive sleep apnea (OSA) and associated comorbidities, and to describe interventions used for OSA.
A retrospective cohort study was performed at Cincinnati Children’s Hospital Medical Center for children with DS born between 2013 and 2019. Data were extracted from the electronic medical record, including demographics, age at PSG, PSG results, and interventions after an abnormal PSG. Statistical analysis included unadjusted bivariate association testing and multivariable logistic regression modeling to investigate associations with OSA severity.
Among 397 patients in the cohort, 59% (n = 235) had a documented PSG and 94% (n = 221) had an abnormal study with 60% (n = 141) demonstrating moderate or severe OSA. There was an inverse relationship between age and OSA severity (P < .001). In a multiple regression model, OSA severity was associated with increased rates of failure to thrive (P < .01), aspiration (P = .02), and laryngomalacia (P < .01). After medical or surgical intervention, 73% of patients experienced the resolution of OSA or an improvement in OSA severity.
In this study of pediatric patients with DS, OSA was identified most frequently in the first year of life. In addition, to prompt evaluation of symptomatic infants, our data support earlier PSG screening for patients requiring neonatal ICU care and those with feeding difficulties, airway abnormalities, and/or pulmonary hypertension given their increased risk for severe OSA.
OSA affects 30% to 70% of children with DS and causes significant morbidity when untreated. Existing literature focuses on older patients with few studies including infants and young children. Current screening guidelines recommend obtaining a sleep study by 4 years of age.
This study reveals that OSA is pervasive among infants with DS, with 98% of sleep studies in patients <1 year of age identifying disease and 59% confirming severe disease. Polysomnographic screening in infancy may facilitate earlier diagnosis and treatment.
Down syndrome (DS) is associated with numerous complications in infancy and early childhood, including sleep-disordered breathing.1–4 Obstructive sleep apnea (OSA) refers to frequent cessation (apnea) or reduction (hypopnea) of airflow during sleep despite the ongoing respiratory effort.5 OSA affects 30% to 70% of pediatric patients with DS, exceeding rates of congenital heart disease, hypothyroidism, and hematologic malignancies, all of which are screened for in the newborn period.6,7 Overnight polysomnography (PSG) is the gold standard for the diagnosis of OSA and other sleep-related disorders.8 Current American Academy of Pediatrics (AAP) health supervision guidelines for children with DS recommend universal PSG screening by 4 years of age.7
The pathophysiology of OSA in DS is multifactorial, with contributions from upper airway anatomic narrowing and collapsibility in the setting of hypotonia. Additionally, medical comorbidities, including hypothyroidism, gastroesophageal reflux, aspiration, recurrent respiratory infections, and seizures contribute to OSA through multiple mechanisms.5,6 Infants are predisposed to airway obstruction and gas exchange abnormalities during sleep because of anatomic and physiologic features, including preferential high-resistance nasal breathing and a lower apneic threshold, and may have more severe OSA compared with older children.9,10
Untreated OSA exacerbates complications that independently impact individuals with DS. Pulmonary hypertension affects approximately one-third of children with DS and is associated with increased mortality.11,12 Upper airway obstruction, transient hypoxemia, and respiratory acidosis in OSA lead to hemodynamic and endovascular alterations that increase pulmonary vascular resistance. These processes occur more readily in patients with DS.13–15 OSA contributes to poor growth through alterations in hormonal signaling and increased energy expenditure from obstructive breathing. The prevalence of failure to thrive (FTT) is at least double in children with OSA compared with the general population.5,16,17 Children with DS are predisposed to FTT because of feeding difficulties, heart disease, and recurrent infections secondary to immune dysfunction.18 Finally, OSA is associated with neurocognitive dysfunction in domains including executive functioning, behavior, and memory, all of which are impaired for patients with DS.19–21
Infants with DS possess overlapping phenotypes that increase the risk of OSA. Despite this vulnerability, existing literature focuses on older children and adults, with few studies including neonates and infants.22,23 To our knowledge, there has not been a comprehensive evaluation of the clinical spectrum of OSA performed in an age-stratified cohort of young patients with DS. With this study, we aimed to evaluate the prevalence and impact of OSA in patients with DS in infancy and early childhood by describing the use of PSG and associated findings, identifying comorbidities associated with OSA, and assessing the utilization of interventions for OSA.
Methods
Patient Population
A retrospective cohort study was performed including all children with DS born between 2013 and 2019 who received care within the Cincinnati Children’s Hospital Medical Center’s (CCHMC) newborn care network. Patients were 0 to 7 years old at the time of data collection. CCHMC, the sole provider of pediatric subspecialty care in the greater Cincinnati area, offers pediatric pulmonology and otolaryngology services and performs PSGs on an inpatient and outpatient basis. Diagnoses of DS were confirmed by karyotype or documentation in the electronic medical record. Data were extracted including demographics, age at PSG(s), PSG parameters and diagnoses, medical comorbidities, and interventions for abnormal PSG results. Data were stored in a secure RedCap database. This study was approved by the CCHMC Institutional Review Board (ID #2020-0308) with a waiver of parental consent.
Clinical Definitions
PSG reports were evaluated by using standard criteria established by the American Academy of Sleep Medicine.24 Patients were diagnosed with OSA if their obstructive index (OI) was ≥1. The OI is defined as the number of obstructive apneas and hypopneas per hour of sleep. OSA was classified as mild (OI of 1 to ≤5), moderate (OI of >5 to ≤10), or severe (OI >10). Additional sleep diagnoses included central sleep apnea, defined as an apnea hypopnea index ≥5, alveolar hypoventilation, defined as end-tidal carbon dioxide values >50 for ≥25% of sleep time, and nonapneic hypoxemia, defined as oxygen saturation of <90% for >2% of sleep time or baseline oxygen saturation <92% not associated with apnea or hypopnea events.8 Interventions were classified as medical, including supplemental oxygen, intranasal steroids, noninvasive positive pressure, and ventilator dependence, or surgical, including tonsillectomy, adenoidectomy, supraglottoplasty, and tracheostomy.
Statistical Analysis
Descriptive data are presented as univariable statistics, including proportions for categorical data and mean/standard deviation for numeric variables. We tested for unadjusted associations between ≥2 categorical variables using χ2 tests or Fisher’s exact tests with 2-sided P values. We used exact tests to assess for differences in comorbidities across OSA severity groups. When investigating differences in age at first sleep study across OSA severity groups, we used 1-way analysis of variance (ANOVA) and Levene’s test to check for violations of equal variances; if heteroscedasticity was present, we performed Welch ANOVA. The accompanying ANOVA model F-statistics and P values are presented.
We modeled OSA severity as an outcome using multinomial logistic regression, with OSA severity having 3 categories; the reference category for odds ratio estimation was “mild.” When appropriate, model parameter estimates, Wald χ2 values, and associated P values are presented; odds ratios with 95% confidence intervals are also provided. We chose independent variables for the regression model on the basis of clinical experience and unadjusted bivariate exploratory data analysis; variables that were not statistically significant were removed from the final model. All analyses were completed by using SAS version 9.4 (SAS Institute, Inc., Cary, NC). The threshold for statistical significance was α <0.05 for 95% confidence.
Results
Demographics and Timing of PSG
We identified 397 patients with DS born between 2013 and 2019 who received care at our institution, and 59% (n = 235) had at least 1 PSG. The remaining 41% of patients (n = 162) did not have a documented PSG; 62% of these patients (n = 100) were aged 4 years or older at the time of data collection and should have undergone PSG per AAP guidelines. The median age at initial PSG was 13 (interquartile range 33) months. There were no significant differences in demographic characteristics between patients who underwent PSG and those who did not (Table 1). In bivariate analyses of demographic variables, worsening OSA severity was associated with male sex and admission to an ICU in the neonatal period (Table 2).
. | Sleep Study, n = 235 (59%) . | No Sleep Study, n = 162 (41%) . | P . | ||
---|---|---|---|---|---|
. | n (%) . | n (%) . | |||
Sex | 1.00 | ||||
Male | 124 | (59) | 86 | (41) | |
Female | 111 | (59) | 76 | (41) | |
Race | .16 | ||||
White | 189 | (61) | 119 | (39) | |
Black | 26 | (55) | 21 | (45) | |
Asian | 6 | (67) | 3 | (33) | |
Multiracial | 6 | (60) | 4 | (40) | |
Other/unknown | 8 | (35) | 15 | (65) | |
Ethnicity | 1.00 | ||||
Non-Hispanic | 223 | (59) | 154 | (41) | |
Hispanic | 12 | (60) | 8 | (40) | |
Mean birth wt, g | 2852 | 2848 | .94 | ||
Mean gestational age, wks | 37.0 | 37.2 | .40 | ||
Neonatal ICU admissiona | 1.00 | ||||
Yes | 166 | (60) | 111 | (40) | |
No | 72 | (60) | 48 | (40) | |
County of residenceb | .20 | ||||
Metropolitan | 191 | (61) | 122 | (39) | |
Rural | 31 | (52) | 29 | (48) | |
Mean maternal age, y | 32.5 | 33.5 | .10 |
. | Sleep Study, n = 235 (59%) . | No Sleep Study, n = 162 (41%) . | P . | ||
---|---|---|---|---|---|
. | n (%) . | n (%) . | |||
Sex | 1.00 | ||||
Male | 124 | (59) | 86 | (41) | |
Female | 111 | (59) | 76 | (41) | |
Race | .16 | ||||
White | 189 | (61) | 119 | (39) | |
Black | 26 | (55) | 21 | (45) | |
Asian | 6 | (67) | 3 | (33) | |
Multiracial | 6 | (60) | 4 | (40) | |
Other/unknown | 8 | (35) | 15 | (65) | |
Ethnicity | 1.00 | ||||
Non-Hispanic | 223 | (59) | 154 | (41) | |
Hispanic | 12 | (60) | 8 | (40) | |
Mean birth wt, g | 2852 | 2848 | .94 | ||
Mean gestational age, wks | 37.0 | 37.2 | .40 | ||
Neonatal ICU admissiona | 1.00 | ||||
Yes | 166 | (60) | 111 | (40) | |
No | 72 | (60) | 48 | (40) | |
County of residenceb | .20 | ||||
Metropolitan | 191 | (61) | 122 | (39) | |
Rural | 31 | (52) | 29 | (48) | |
Mean maternal age, y | 32.5 | 33.5 | .10 |
Admission to a NICU or CICU within the first 28 days of life.
County type was designated according to the United States Office of Management and Budget definitions. A metropolitan area is defined as containing a core urban area of 50 000 or more population. This was used as a surrogate indicator of proximity to the medical center.
. | OSA Severity, n (%) . | . | |||
---|---|---|---|---|---|
. | No OSA . | Mild . | Moderate . | Severe . | P . |
Sex | <.01 | ||||
Male | 6 (5) | 37 (30) | 23 (19) | 57 (46) | |
Female | 5 (5) | 43 (39) | 32 (29) | 29 (27) | |
Race | .65 | ||||
White | 9 (5) | 65 (35) | 45 (24) | 67 (36) | |
Black | 0 (0) | 7 (27) | 7 (27) | 12 (46) | |
Asian | 0 (0) | 1 (33) | 2 (67) | 0 (0) | |
Multiracial | 0 (0) | 4 (67) | 0 (0) | 2 (33) | |
Other/unknown | 2 (25) | 3 (38) | 1 (12) | 2 (25) | |
Ethnicity | .29 | ||||
Non-Hispanic | 9 (4) | 74 (34) | 54 (25) | 83 (37) | |
Hispanic | 2 (17) | 6 (50) | 1 (8) | 3 (25) | |
Mean birth wt, g | 2896 | 2896 | 2894 | 2782 | .67 |
Mean gestational age, wks | 37.3 | 37.2 | 37.1 | 36.7 | .52 |
Neonatal ICU admissiona | .03 | ||||
Yes | 5 (3) | 47 (30) | 39 (24) | 65 (41) | |
No | 4 (5) | 34 (45) | 16 (21) | 20 (27) |
. | OSA Severity, n (%) . | . | |||
---|---|---|---|---|---|
. | No OSA . | Mild . | Moderate . | Severe . | P . |
Sex | <.01 | ||||
Male | 6 (5) | 37 (30) | 23 (19) | 57 (46) | |
Female | 5 (5) | 43 (39) | 32 (29) | 29 (27) | |
Race | .65 | ||||
White | 9 (5) | 65 (35) | 45 (24) | 67 (36) | |
Black | 0 (0) | 7 (27) | 7 (27) | 12 (46) | |
Asian | 0 (0) | 1 (33) | 2 (67) | 0 (0) | |
Multiracial | 0 (0) | 4 (67) | 0 (0) | 2 (33) | |
Other/unknown | 2 (25) | 3 (38) | 1 (12) | 2 (25) | |
Ethnicity | .29 | ||||
Non-Hispanic | 9 (4) | 74 (34) | 54 (25) | 83 (37) | |
Hispanic | 2 (17) | 6 (50) | 1 (8) | 3 (25) | |
Mean birth wt, g | 2896 | 2896 | 2894 | 2782 | .67 |
Mean gestational age, wks | 37.3 | 37.2 | 37.1 | 36.7 | .52 |
Neonatal ICU admissiona | .03 | ||||
Yes | 5 (3) | 47 (30) | 39 (24) | 65 (41) | |
No | 4 (5) | 34 (45) | 16 (21) | 20 (27) |
Admission to a NICU or CICU within the first 28 days of life.
PSG Indications
Most patients (82%, n = 193) had a PSG because of the presence of symptoms, including snoring, apnea, stridor, or the inability to wean supplemental oxygen; the remainder (18%, n = 42) underwent asymptomatic screening. Symptomatic patients were younger at the time of initial PSG (16 months vs 37 months, P <.001) and had a higher OI (15 vs 6.3, P <.001) compared with patients undergoing screening PSG; however, the overall rates of OSA between symptomatic and asymptomatic groups were not significantly different (96% vs 93% with any degree of OSA, P = .2).
OSA Severity and Additional Sleep Diagnoses
Among patients who had a sleep study, 94% (n = 221) had OSA with an average OI of 13.5 (15.5, range 1–108). After the evaluation of each patient’s first PSG, 34% (n = 80) were classified as mild OSA, 23% (n = 55) moderate, and 37% (n = 86) severe. The remaining 6% of patients had no OSA. Severe OSA was diagnosed most frequently in patients ≤6 months of age (Table 3), and there was a significant inverse relationship between age and OSA severity (P < .001, Fig 1). Other PSG diagnoses in the cohort included alveolar hypoventilation (53%), nonapneic hypoxemia (58%), and central sleep apnea (43%). OSA severity was significantly associated with the presence of an additional sleep pathology, with 70% of patients with severe OSA having an additional diagnosis compared with 26% of patients with mild OSA (P <.01).
. | Severe . | Moderate . | Mild . | No OSA . | Total . | . | |||||
---|---|---|---|---|---|---|---|---|---|---|---|
. | n . | % . | n . | % . | n . | % . | n . | % . | n . | % . | Mean OI . |
0–6 mo | 55 | 66 | 19 | 23 | 9 | 11 | 0 | 0 | 83 | 36 | 20.0 |
7–12 mo | 12 | 38 | 7 | 22 | 11 | 34 | 2 | 6 | 32 | 14 | 9.9 |
13–23 mo | 7 | 32 | 5 | 23 | 9 | 41 | 1 | 4 | 22 | 9 | 15 |
2–3 y | 1 | 5 | 8 | 38 | 13 | 61 | 0 | 0 | 22 | 9 | 8.2 |
3–4 y | 8 | 28 | 8 | 28 | 11 | 38 | 2 | 6 | 29 | 13 | 10 |
≥4 y | 2 | 5 | 8 | 19 | 27 | 64 | 5 | 12 | 42 | 18 | 3.8 |
. | Severe . | Moderate . | Mild . | No OSA . | Total . | . | |||||
---|---|---|---|---|---|---|---|---|---|---|---|
. | n . | % . | n . | % . | n . | % . | n . | % . | n . | % . | Mean OI . |
0–6 mo | 55 | 66 | 19 | 23 | 9 | 11 | 0 | 0 | 83 | 36 | 20.0 |
7–12 mo | 12 | 38 | 7 | 22 | 11 | 34 | 2 | 6 | 32 | 14 | 9.9 |
13–23 mo | 7 | 32 | 5 | 23 | 9 | 41 | 1 | 4 | 22 | 9 | 15 |
2–3 y | 1 | 5 | 8 | 38 | 13 | 61 | 0 | 0 | 22 | 9 | 8.2 |
3–4 y | 8 | 28 | 8 | 28 | 11 | 38 | 2 | 6 | 29 | 13 | 10 |
≥4 y | 2 | 5 | 8 | 19 | 27 | 64 | 5 | 12 | 42 | 18 | 3.8 |
Neonatal ICU Admission
There were 277 patients (70% of the cohort) who required admission to an ICU in the first 28 days of life. The most common indications for admission were hypoxia/respiratory distress (58%), feeding difficulties (28%), congenital heart disease (16%), noncardiac congenital anomalies (16%), and prematurity (16%). Among neonates admitted to the ICU, 12% (n = 34) underwent PSG during hospitalization. All studies performed during ICU admission were diagnostic of OSA and 74% (n = 25) showed severe disease, with an average OI of 23.8. Overall, 88 patients (32%) who required neonatal ICU care had a PSG during the first year of life and 99% (n = 87) of these studies revealed OSA, 59% of which was severe. Neonatal ICU admission was associated with worse OSA severity, with 41% (n = 65) of patients requiring intensive care having severe OSA on their initial PSG compared with 27% (n = 20) of patients who required typical newborn care (P = .03).
Medical Comorbidities
There were high rates of feeding difficulties and need for artificial feeding support in the cohort. Among 178 patients who had a video swallow study, 39% (n = 69) demonstrated aspiration. Dysphagia was diagnosed in 45% (n = 178) of the cohort by video swallow study or clinical assessment. Overall, 32% (n = 127) of the cohort required a gastrostomy tube or home nasogastric tube feeds. FTT was diagnosed in 17% (n = 68) of patients. Airway abnormalities were identified in 43% (n = 171) of the cohort, including laryngomalacia (n = 126), tracheomalacia (n = 76), subglottic stenosis (n = 34), laryngeal cleft (n = 13), and tracheal rings (n = 12). Patients with severe OSA were more likely to undergo an airway evaluation, including microlaryngoscopy and bronchoscopy, compared with those with mild OSA (77% vs 36%, P <.01). All patients in the cohort had a documented echocardiogram; 85% (n = 337) of patients had a structural heart defect and 26% (n = 103) had pulmonary hypertension.
Association of Comorbidities With OSA Severity
Bivariate testing was used to evaluate unadjusted associations between OSA severity and health conditions prevalent in young children with DS (Table 4). Multinomial logistic regression modeling was then used to evaluate the effect of significant variables on OSA severity while controlling for patient age. Given the high rates of OSA in the cohort, mild OSA was used as the reference category. Aspiration, FTT, and laryngomalacia remained significant predictors of OSA severity in this model (Fig 2). Pulmonary hypertension was also associated with increased odds of moderate or severe OSA and was highly correlated with OSA severity in our bivariate analysis (P <.01) but did not reach statistical significance in the regression model after adjusting for age.
. | Severe . | Moderate . | Mild . | No OSA . | . | ||||
---|---|---|---|---|---|---|---|---|---|
. | n . | % . | n . | % . | n . | % . | n . | % . | P . |
Hypothyroidism | 28 | 33 | 18 | 33 | 27 | 34 | 4 | 36 | 1.00 |
Failure to thrive | 33 | 38 | 8 | 15 | 26 | 33 | 1 | 9 | <.01 |
Seizures | 6 | 7 | 3 | 5 | 7 | 9 | 1 | 9 | .84 |
Aspiration | 31 | 36 | 15 | 27 | 9 | 11 | 3 | 27 | <.001 |
G-tube or NG feeds | 34 | 40 | 19 | 34 | 22 | 28 | 5 | 45 | <.05 |
Laryngomalacia | 57 | 66 | 18 | 33 | 25 | 31 | 4 | 36 | <.001 |
Tracheomalacia | 29 | 34 | 8 | 15 | 15 | 19 | 4 | 36 | <.05 |
Structural heart disease | 77 | 38 | 48 | 23 | 67 | 33 | 11 | 5 | .61 |
Pulmonary hypertension | 31 | 47 | 20 | 30 | 13 | 20 | 2 | 3 | <.01 |
. | Severe . | Moderate . | Mild . | No OSA . | . | ||||
---|---|---|---|---|---|---|---|---|---|
. | n . | % . | n . | % . | n . | % . | n . | % . | P . |
Hypothyroidism | 28 | 33 | 18 | 33 | 27 | 34 | 4 | 36 | 1.00 |
Failure to thrive | 33 | 38 | 8 | 15 | 26 | 33 | 1 | 9 | <.01 |
Seizures | 6 | 7 | 3 | 5 | 7 | 9 | 1 | 9 | .84 |
Aspiration | 31 | 36 | 15 | 27 | 9 | 11 | 3 | 27 | <.001 |
G-tube or NG feeds | 34 | 40 | 19 | 34 | 22 | 28 | 5 | 45 | <.05 |
Laryngomalacia | 57 | 66 | 18 | 33 | 25 | 31 | 4 | 36 | <.001 |
Tracheomalacia | 29 | 34 | 8 | 15 | 15 | 19 | 4 | 36 | <.05 |
Structural heart disease | 77 | 38 | 48 | 23 | 67 | 33 | 11 | 5 | .61 |
Pulmonary hypertension | 31 | 47 | 20 | 30 | 13 | 20 | 2 | 3 | <.01 |
Interventions for OSA
For patients in the cohort with a diagnosis of OSA, 93% (n = 218) required 1 or more medical or surgical interventions (Table 5). Interventions varied by patient age, OSA severity, and associated comorbidities (eg, structural airway abnormalities). Patients with moderate or severe OSA were more likely to require multiple interventions than patients with mild OSA (88% vs 85% vs 64%, P <.01). Additionally, patients with severe OSA underwent surgical intervention or initiation of positive pressure at a younger age compared with those with milder disease. There were 172 patients (73% of those with an initial PSG) who had at least 1 follow-up PSG. Among these patients, 28% (n = 48) had resolution of OSA and an additional 45% (n = 78) had improvement by ≥1 severity category after intervention(s) (Fig 3). For this analysis, OSA severity was classified by the patient’s worst OI and resolution was defined as an OI <2 per the Childhood Adenotonsillectomy Trial.25
Any Intervention, n = 218 . | n . | 93% . | Median Age, Mo . |
---|---|---|---|
Medical | |||
Supplemental oxygen | 112 | 51 | 4 |
Intranasal steroids | 40 | 17 | 24 |
CPAP/BiPAP | 30 | 13 | 48 |
Ventilator dependence | 5 | 2 | 3 |
Surgical | |||
Adenoidectomy | 159 | 70 | 24 |
Tonsillectomy | 146 | 64 | 36 |
Supraglottoplasty | 59 | 27 | 7 |
Othera | 16 | 7 | 42 |
Tracheostomy | 7 | 3 | 2 |
Multiple interventions | 178 | 79 | Not applicable |
Any Intervention, n = 218 . | n . | 93% . | Median Age, Mo . |
---|---|---|---|
Medical | |||
Supplemental oxygen | 112 | 51 | 4 |
Intranasal steroids | 40 | 17 | 24 |
CPAP/BiPAP | 30 | 13 | 48 |
Ventilator dependence | 5 | 2 | 3 |
Surgical | |||
Adenoidectomy | 159 | 70 | 24 |
Tonsillectomy | 146 | 64 | 36 |
Supraglottoplasty | 59 | 27 | 7 |
Othera | 16 | 7 | 42 |
Tracheostomy | 7 | 3 | 2 |
Multiple interventions | 178 | 79 | Not applicable |
Other surgical interventions included nasal turbinate resection, partial glossectomy, and cricoid mass resection.
Discussion
In this study of a large cohort of infants and young children with DS, >90% of patients who underwent PSG were diagnosed with OSA. The most significant findings were in infants (≤1 year old), with 98% of this age group having an abnormal PSG and 58% carrying a diagnosis of severe OSA, with a mean OI of 19. In contrast, patients who underwent PSG at age ≥4 years mostly had mild OSA, with a mean OI of 3.8. Younger age was a highly significant predictor of severe OSA after controlling for other clinical and demographic factors (Fig 2). This study is the first to establish that infants with DS are disproportionately affected by severe OSA compared with older age groups. These data suggest that PSGs obtained in infancy would detect a large percentage of severe OSA, allowing for prompt intervention and possibly improved long-term outcomes.
Our multiple regression model yielded significant associations between moderate to severe OSA and FTT, aspiration, and laryngomalacia. Goffinski et al also found a relationship between OSA and gastrointestinal anomalies in a cohort of patients ≤6 months of age with DS.23 Breathing and feeding are interrelated processes in all infants, and infants with DS have particularly high rates of feeding dysfunction, including dysphagia and aspiration.18 Aspiration may lead to inflammation of the upper airway and blunting of pharyngo-laryngeal reflexes involved in swallowing, leading to the pooling of secretions in the hypopharynx with subsequent obstruction. On the basis of our data, we suggest that all patients with DS and feeding difficulties or upper airway abnormalities undergo PSG when these problems are identified. Although our model did not reveal a significant relationship between pulmonary hypertension and OSA severity, there is an abundance of clinical and physiologic evidence supporting the interplay between these 2 disorders. PSG should remain part of the standard evaluation for children with DS and new or worsening pulmonary hypertension.11–13,15
Among term infants with DS without major structural anomalies, respiratory distress and hypoxemia are frequent indications for admission to the NICU.4,26,27 Our data suggest that many newborns with OSA are being discharged from the ICU without evaluation. For patients with DS who require ICU care in the neonatal period, obtaining an inpatient PSG before discharge or referring for an outpatient PSG shortly thereafter, depending on institutional capabilities, should be strongly considered given the high rates of severe OSA in this subpopulation.
Notably, 93% of asymptomatic patients in our cohort met polysomnographic criteria for OSA. Previous studies have revealed that clinical assessment, parental report of symptoms, and questionnaire-based screening tools do not accurately predict the presence of OSA in children with DS.28,29 Modalities such as overnight oximetry and capnography have similarly poor diagnostic sensitivity.30 Our data reinforce the role of PSG as an essential diagnostic tool as OSA is difficult to predict on the basis of clinical assessment alone and bolster the AAP recommendation for universal PSG screening for children with DS.
Overall, 41% of the cohort did not have a documented PSG, including 33% of patients who were aged ≥4 years at the time of data collection. Studies suggest that a lack of physician and caregiver awareness, resource limitation, and medical/behavioral comorbidities contribute to nonadherence.31–33 Pediatric sleep studies are not universally available in the United States; however, this acts as a barrier to screening regardless of the age at which testing is recommended. Interestingly, all patients in our cohort had a documented echocardiogram, a study that is equivalent in cost and accessibility to PSG in our region. This suggests that, although access is a barrier to OSA evaluation for some patients, it was not the driving factor in our population. Earlier PSG acquisition may have the added benefit of improved screening rates given frequent contact with the health care system in infancy, increased scheduling flexibility because infant sleep studies are not necessarily performed overnight, and fewer behavioral barriers to study completion in infants versus older children.
Race and ethnicity were included in our evaluation of PSG acquisition and OSA severity because existing literature reveals poorer health outcomes for nonwhite patients with DS, as well as disparities in OSA screening and severity.32,34–36 Although there appeared to be a trend toward lower rates of PSG obtainment and higher rates of severe OSA for Black patients in the cohort, neither reached statistical significance. Additionally, there was no difference in age at first PSG among racial or ethnic groups.
Most patients required multiple interventions for OSA over time, which is likely attributable to the multifactorial pathogenesis of upper airway obstruction in DS. Previous studies reveal that patients with DS have less polysomnographic and symptomatic improvement after adenotonsillectomy or supraglottoplasty compared with controls and are more likely to require ongoing respiratory support postoperatively.37–40 Although rates of complete OSA resolution were low, 73% of patients demonstrated improvement by at least 1 severity category after intervention. Earlier diagnosis and subsequent treatment, although not always curative, have the potential to mitigate transient hypoxemia, which is known to worsen neurocognitive impairment and pulmonary hypertension in patients with DS.20,41
This study has limitations that we acknowledge. Data collection was limited to the CCHMC electronic medical record, and PSGs performed at outside institutions may have been inconsistently documented. It is possible that patients referred for PSG were more likely to have OSA, leading to selection bias and the overestimation of prevalence; however, this is unlikely given asymptomatic disease and difficulty detecting OSA solely on the basis of clinical suspicion, as previously discussed. If we assume that patients without a sleep study had no OSA, prevalence remains high at 55% for OSA overall and 36% for severe OSA. Although the cohort was composed of patients from a single center, nearly all children with DS in the region are born at an affiliated hospital or receive subspecialty services through our institution; therefore, we were able to capture a generalizable population-based sample. Because of the retrospective nature of the study, causal relationships between OSA and the comorbidities of interest could not be established.
Finally, there is a paucity of normative values for the interpretation of PSG in neonates. A recent study by Daftary et al found that, for a small cohort of healthy term newborns, the average apnea hypopnea index was 14.9; however, obstructive apnea was uncommon in these patients, and desaturations were mild and self-resolving.42 Further study is needed to establish baseline PSG parameters specific to infants with DS because this population demonstrates sleep architecture and pathology that is distinct from nonsyndromic patients5,6,43
Conclusion
OSA is highly prevalent among infants and young children with DS and is an early, modifiable complication contributing to long-term morbidity. Current AAP guidelines recommend a sleep study be performed before 4 years of age. The pervasiveness of severe disease requiring intervention in infants suggests that earlier testing may be warranted, particularly for patients requiring neonatal intensive care and those with feeding difficulties, upper airway abnormalities, and/or pulmonary hypertension. Prospective study is needed to understand the true prevalence of OSA in all infants with DS, identify optimal intervention(s), and determine how earlier diagnosis and treatment impact rates of adverse outcomes.
Dr Seither designed the data collection instruments, collected the data, participated in data analysis, drafted the initial manuscript, and reviewed and revised the manuscript; Mr Helm performed the data analysis and reviewed and revised the manuscript; Drs Heubi and Swarr contributed to the study design and reviewed the manuscript; Dr Suhrie conceptualized and designed the study, supervised data collection, and critically reviewed the manuscript for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no potential conflicts of interest relevant to this article to disclose.
Comments