Down syndrome (DS) is the most frequently occurring chromosomal disorder and leading cause of intellectual disability in the United States. With an incidence of 1/700 live births, an estimated 6,000 infants with DS are born in the US each year [1]. Children with DS are at a higher risk for neonatal complications and medical comorbidities as compared to their typically developing peers. According to the most recent report and practice guidelines for monitoring the health of children with DS published by the American Academy of Pediatrics (AAP) in 2011, these comorbidities include the occurrence of congenital heart defects (40-50%), gastrointestinal (GI) malformations (12%), sleep apnea (50-75%), thyroid disease (4-18%), celiac disease (5%), autism (1%), childhood leukemia (1%), transient myeloproliferative disorder (10%) and seizures (1-13%) [2]. Current AAP Guidelines recommend testing with an echocardiogram at birth, audiologic and ophthalmologic evaluations, thyroid labs, a swallow study if a feeding problem is present and a sleep study at four years of age. Although increased comorbidity incidence has been identified, only 5-10% of the patient population are reported to be in compliance with these guidelines [3].

Few studies have investigated comorbidities in children with DS through application of the AAP Guidelines [4-6]. These investigations are significant in establishing a baseline to better understand this population. However, there are some limitations that include small participant numbers, data collection spanning many years, disparities in subject age range, lack of ethnic or race information and limited reliability of data systems for clinical information [7-15]. These challenges in the literature impact the reported frequency of comorbidities and limit our understanding of screening and identification of medical conditions for children with DS. This is especially evident with autoimmune thyroid conditions (7-50%), seizures (1-10%), infantile spasms (3-15%), autism (1-7%) and obstructive sleep apnea (OSA) (30-60%), where reported frequencies vary widely [7-10]. An improved understanding of the complications and comorbidities impacting children with DS are the next steps toward improving disease management in this population.

Until recently, published articles have mostly relied on surveys and parental report with limited data collection from DS clinics [4,6,11]. Parental reports are limited by a reliance on the parent’s ability to recall medical histories accurately. Existing literature regarding the validity of parent reported comorbidities is sparse and report large discrepancies in accuracy in regards to different medical conditions [16]. The design and implementation of the Anna and John J Sie Centre for Down Syndrome (SCDS) provided a unique opportunity to gather medical information and population data with each clinic visit to better understand patients’ needs based on medical chart documentation. Developed in 2011, the SCDS is a multidisciplinary specialty clinic providing evaluation, consultation and therapies for children and young adults diagnosed with DS from the prenatal phase of life up to age 21. The closest DS clinics are located over 600 and 1,000 miles away; therefore, along with the large patient population, the clinic provides a more population based catchment model. This unique location also offers services to the medically understudied and underserved Hispanic (21.2%) and immigrant populations in Colorado. The SCDS developed one of the largest current databases of medical comorbidities in children with DS in the US comprised of over 1,100 unique pediatric patients evaluated by a single developmental paediatrician and follow-up care including a nurse practitioner at one institution.

Although the association of comorbidities with DS is well known, details such as the actual incidence and nature of these medical complications and overall impact of the AAP Guidelines in the DS population are lacking. These findings provide a foundation for improving screening for medical conditions, potentially allowing earlier interventions and improvements in quality of life, morbidity and mortality in this population. Therefore, the objective of this article was to better understand and quantify the unique phenotypic clinical pattern of comorbidities and medical complications in children and adolescents with DS using a large multi-age and ethnic sample applying current AAP Guidelines testing practices.

Design

This study investigates neonatal complications and incidence of medical comorbidities in 1,108 children with DS within a recent fiveyear period utilizing a retrospective collection of clinic medical data. All participants were treated in a multidisciplinary DS specialty clinic at a single, large, pediatric academic medical hospital. Patients were identified and data were extracted from the SCDS Clinical REDCap (Research Electronic Data Capture) [17] Database populated by a clinic form and electronic medical chart review. Data collected included prenatal and birth history and complications, comorbidities, testing results and any normal or abnormal findings. The study was approved by the Institutional Review Board of Children’s Hospital Colorado, University of Colorado School of Medicine, which determined that the need for informed consent was exempted.

Participants/subjects

All participants (male=602; mean age at initial contact: 5.72 years, SD ± 5.51) met inclusion criteria of having attended at least one SCDS clinic appointment between January 1, 2011 to December 1, 2016, age one day to 21 years at time of receiving care and a diagnosis of DS. Demographic data for the 1,108 children are displayed in Table 1. Colorado Department of Public Health and Environment comparative data of all children born in the state from 2000-2013 are also reported. Data results indicate that our study captures approximately 50.3% of the state of Colorado’s population of children with DS and provides support of a population based study. The high representation of Hispanic and Latino SCDS patients (28.6%) reflect the demographics in the community, as well as the outreach efforts of the clinic. The distribution of our population with DS (Trisomy 21 nondisjunction=95.8%; Translocation DS=2.6%; Mosaic DS=1.6%) is similar to data reported in the genetic literature [18].

Table 1: Colorado State and clinic patient demographics of children with Down syndrome.

Measures

Neonatal and comorbidity information were derived from the SCDS Clinic Form completed by the medical director and program director and review of the electronic medical record. Data collected from this electronic form were analysed to assess the investigation’s main outcome measures: (1) Birth weight and prematurity, (2) maternal age, (3) neonatal health complications and intensive care unit admittance, (4) abnormal new-born hearing screen results, (5) medical comorbidities and complications: Gastrointestinal, ear-nose-andthroat, audiology, autoimmune, neurologic, autism and neurologic diagnoses, cardiac defects and sleep study abnormalities and (6) association between ethnicity and medical comorbidities.

Procedure

Patients were identified from the SCDS Clinical REDCap Database, which populates the medical history of all patients seen in the SCDS since its implementation in 2011. Study data were collected from the SCDS Clinic Visit Form completed by the medical director and program director for each patient visit and using the electronic medical record. Additional medical information from external institutions was requested when applicable.

Following a patient’s first SCDS clinic visit, the medical director or program director completed the Clinic Visit Form relating to the child’s evaluation and prior testing, diagnoses and treatment. Specific information about patient demographic data (e.g. race, ethnicity, age at appointment, primary language of parents or guardians, sex), prenatal and birth history and complications, comorbidities, testing and clinical recommendations were included. Additionally, a separate Clinic Visit Form was completed for every child at each subsequent time the child’s history had changed (e.g. new diagnosis, test) at time of an encounter.

Diagnostic billing codes and EMR review identified diagnoses. When specific diagnoses were not available, diagnoses were determined using clinically accepted criteria. Children were classified as having aspiration from abnormal swallow study results. A thyroid diagnosis required an abnormal thyroid test that required medical treatment. Ophthalmologic abnormalities were defined as abnormalities noted by an ophthalmologist. Gastroesophageal reflux disease (GERD) was recorded if the medical records documented the diagnosis by a medical doctor. Feeding problems in the neonatal period were defined as requiring nasal gastric tube feeding or G-tube placement. Celiac disease screening labs were performed on all children evaluated in clinic over three to four years of age instead of relying solely on symptomatic testing.

Patient demographics, concomitant medical problems and procedures and history of neonatal complications were queried from the database. Patients from the SCDS Clinic Database were assigned unique study IDs and de-identified prior to being uploaded into the study database. Accuracy of data collection was verified by the research assistant through review of the REDCap database.

Descriptive statistics and Chi-squared test were performed on demographic and clinical characteristics in the data set. Results are presented as mean ± standard deviation and range or percentage where appropriate. Because we tested ethnic differences separately for each of the medical comorbidities and complications, we used the Benjamini and Hochberg false discovery rate (FDR) to control for potential Type I errors [19]. T-tests were run to determine the significant associations between ethnicity and maternal age. Data were analysed by the statistical package SPSS 24.0. All analyses were conducted at α-level 0.05 significance. Children were omitted in an analysis if the variable of investigation had an “unknown” or “missing data” value.

Neonatal complications

Overall, 28.2% (n=283/1,002) of patients seen in SCDS were born prematurely, defined as less than 37 weeks, in contrast to 10% of all US births [20,21]. Birth weight ranged from 1999 g and under (10.1%, n=101/1002), 2000-2499 g (18.2%, n=182), 2500-2999 g (32.8%, n=329) and 3000 or more grams (38.9%, n=390) (Table 2).

Table 2: Descriptive statistics of patient demographics.

Maternal age at birth is displayed in Table 3. There was a statistically significant difference in maternal age between Hispanic/Latina women (M=31.52 years, SD=7.99) and non-Hispanic/Latina women (M=33.49 years, SD=6.54); t (979)=-4.02, p<0.000.

Table 3: Maternal age.

The majority of babies with DS experienced neonatal complications after birth as gathered by medical chart review. Babies most often required neonatal intensive care unit (NICU) stay (n=777, 70.6%) (Table 4). Feeding problems, defined as requiring nasogastric tube or gastronomy tube (G-tube), were also prevalent (n=534, 48.2%). Newborn hearing screen data was only available on 714 children.

Table 4: Neonatal diagnoses and complications.

Comorbidities and medical complications

Descriptive statistics were run on the 1,108 patients to determine the frequency of comorbidities occurring in children who were treated at the SCDS (Table 5). Heart defects included atrial septal defect (n=329, 29.7%), ventricular septal defect (n=230, 20.8%), atrioventricular canal (n=154, 13.9%), patent ductus arteriosus requiring surgical repair (n=137, 12.4%), aberrant subclavian artery (n=22, 2.0%), coarctation (n=19, 1.6%), tetralogy of fallot (n=16, 1.4%), vascular ring (n=3, 0.3%) and other (n=160, 14.4%). The other heart defect category was comprised of open-ended diagnoses such as valve defects (bicuspid, dysplastic, stenosis, etc.) and aortic arch hypoplasia.

Table 5: Comorbidities and medical complications of children treated in the SCDS Clinic.

Chi-square tests for association were conducted between ethnicity and medical comorbidities listed in Table 5. There were no statistically significant associations (p<0.05) between ethnicity and comorbidities, except for thyroid abnormalities and GERD. There was a statistically significant finding for thyroid abnormalities and ethnicity, χ²(1)=9.744, p=0.002 with Hispanics/Latinos more often diagnosed (24.2%) than others (15.2%). GERD was also statistically significant, χ²(1)=7.912, p=0.005 with non-Hispanic/Latinos (64.8%) more likely to be diagnosed than Hispanics/Latinos (53.1%). There were no other statistically significant associations (p<0.05) between ethnicity and the medical comorbidities and medical complications analysed when performing the Benjamini and Hochberg FDR control model at an alpha level of 5%.

This study describes the comorbidities affecting individuals with DS in a more accurate way by studying a larger population than previously defined. Results support our understanding of neonatal history and comorbidities uncovered through rigorous application of the AAP DS Guidelines and utilization of a comprehensive clinic database. Despite the large number of retrospective and prospective articles regarding comorbidities in children with DS, these studies are limited to a small number of clinical investigations, with no more than 763 patients and spanning over 25 years [4]. This study reflects one of the largest clinical investigations at a single pediatric hospital over a five-year time period, seen at initial visit by the same developmental paediatrician. In addition, this investigation uses a large multi-age and ethnic sample and current AAP Guidelines testing practices.

Neonatal complications

Our patients experienced a high rate of neonatal complications with 70.6% requiring NICU stay as new-borns, indicating the importance of appropriate medical readiness and intervention from birth to discharge [4,22]. Infants required oxygen in over 60% of admits and 60% required phototherapy. These numbers are higher in comparison to the largest NICU dataset for DS published from the United Kingdom (n=725, NICU admission rate=46%; require oxygen=31%). The high incidence of comorbidities in this population highlights the need for paediatrician and family practice physician awareness regarding the increased risk for neonates with DS. Compliance with the AAP Guidelines screenings for neonates with DS, such as ECHO and CBC testing, are also needed to identify comorbidities early.

Comorbidities

Evidence of the high incidence of comorbidities in children with DS indicates the medical complexity of these patients. Considering the young age of this study population (60%

Cardiac defects: Our cardiac comorbidity data supports the high incidence reported in the literature. The high referral rate to cardiology (33%) emphasizes the significant cardiac care that was lost to follow-up prior to a patient’s SCDS clinic visit. An echocardiogram at birth was lacking in patients prompting the SCDS to develop an Infant Clinic to address this critical need. One exception to the literature, AV canal (13.9%), is lower in our population than the previously identified 25-37%. Thirty-one percent of our population required surgical intervention for cardiac abnormalities, which is one of the few accounts reported in the literature but often a frequent question from parents [12-14,23].

GI diagnosis: GI anomalies (12.1%) and gastroesophageal reflux (38%) are consistent with the published prevalence reported [5]. Esophageal stricture (1.6%), which has been reported to be related to GERD, is often not described in past studies and another silent complication that has a significant medical impact including feeding problems, risk of precancerous lesions requiring multiple surgeries or complicated long term management. Earlier diagnosis would allow for improved outcome.

Autoimmune diagnosis: Thyroid abnormalities (29.1%), hypothyroid (27.3%) and celiac disease (5.0 %) results may be an underestimation, since a large percentage of our clinic population (26.5%) is under four years of age. We project that patients with celiac disease will be over the 5% range due to our younger age bias, as well as the on-going lifetime risk. Most of our celiac patients were asymptomatic at diagnostic testing. Though not currently in the guidelines, we believe that celiac disease is a silent comorbidity requiring testing of all of our symptomatic and asymptomatic patients. The majority of thyroid comorbidities, except for a few with hyperthyroidism, were identified by screenings without prior symptoms, emphasizing the importance of adhering to the AAP DS Guidelines. This increase in diagnosis of thyroid abnormalities in Hispanics/Latinos children may reflect a lack of ordering of thyroid labs previous to the child’s SCDS clinic evaluation.

Aspiration: In the first few years of the clinic, a significant number of abnormal swallow studies with aspiration (12.2%) were noted in our population data. The 2011 AAP DS Guidelines also recommended increased consideration of obtaining feeding assessments in children with DS [2]. These two issues led to the following: (1) Increase in ordering of swallow studies, (2) starting a Feeding Clinic and (3) initiating subsequent feeding research. Our research led to the finding that 91% of patients with abnormal swallow studies aspirated silently without symptoms [24]. Follow-up studies are in process to better define at risk infants.

Leukemia: Leukemia (1.5%) incidence is consistent with data in the literature, whereas transient myelodysplasia (1.6%) is significantly lower than the 10% reported [2,4,25]. We are unsure of the reason regarding the difference in myelodysplasia rates but suspect that a part of this variation may be due to: (1) This large significant clinic population, (2) lack of a complete blood count at birth in older patients and (3) potential unintentional bias in the literature due to reports from oncology centres [26].

Sleep: Of the patients with completed sleep studies, 71.1% had abnormal results with 66.1% experiencing OSA. These results demonstrate one of the largest cohorts reported and potentially a more accurate incidence rate for this population. The number of polysomnography sleep studies (PSG) ordered in the SCDS has increased since 2011 to maintain compliance with AAP DS Guidelines testing recommendations at age four. Increased PSGs have improved identification of patients requiring OSA and the need for organized treatment, prompting the development of a corresponding DS sleep clinic staffed by a pulmonologist, ENT and DS provider. These findings led to patients requiring tonsillectomy and adenoidectomy estimates, treatment with continuous positive airway pressure (CPAP), the development of a corresponding desensitization program for children with DS to increase compliance and initiating research on high flow nasal oxygen for children with DS.

Ophthalmologic abnormalities: Ophthalmologic abnormalities (65.2%) and the need for surgical intervention (nasolacrimal duct obstruction surgery=8.0%, strabismus surgery=5.3%) reinforce the importance of yearly appointments. Primary care physicians and clinics should realize the importance of assessment prior to six months and yearly ophthalmologic follow-up with this population [27]. Incidences of cataracts are increased (3.9%) than what has been previously reported [28].

Audiology: Our results regarding PE tube (42.2%) and hearing aid (9.0%) use provides support of hearing problems in children and young adults with DS. Abnormal newborn hearing screen results in 22.8% is consistent with Utah DS neonate data [2]. The high rate of audiologic and hearing anomalies in our patients reinforces the DS Guideline recommendations for testing every 6 months until three years of age and then annually [29]. Our highest noncompliant rate of all guideline recommendations is audiology at 74%, which is disconcerting. A recent published paper from our clinic and audiology department at Children’s Hospital Colorado emphasizes the clinical importance of the recommended auditory guidelines. In 2013, after increasing referrals from our clinic to audiology, 24.9% of 308 children with DS were identified with permanent hearing loss at the average age of 6 years [30].

Ethnicity: The SCDS patient diversity reflects a higher percentage of Hispanic and Latino families with children with DS than in the literature. The Hispanic/Latino population represents a significant ethnic proportion of the DS population and emphasizes the need for providing education and outreach to families and to practitioners serving these families [31]. The SCDS’s targeted outreach and collaboration with the Hispanic and Latino community DS group has increased awareness of clinical screening and needs. As a result of the outreach to Hispanic families with children with DS, the clinic: (1) has seen an increase of Hispanic and Latino patients compared to other investigations, more accurately reflecting the distribution within the state of Colorado (over 20%), (2) emphasized the importance of appropriate outreach for this group and (3) provided an opportunity to have this underserved and understudied population participate in DS clinical research.

There were limitations to our study that are inherent to a clinical review including a lack of control of treatment parameters that may have impacted outcomes and variability in historical medical records and testing evaluation reports available. Although all medical history, testing and outcomes were reviewed by a board-certified developmental paediatrician and certified paediatric nurse practitioner with expertise in treating and working with children with DS, some inconsistencies in the available data details were inevitable. One might also speculate that children receiving services from a single pediatric hospital can potentially create a population bias. Our sample includes a disproportionately lower percent of Black/African American and higher percent of Hispanic/Latino children and young adults as compared to other previous studies, which may not appropriately address the specific needs of these demographics [31]. This study was designed to reduce limitations in the existing literature through one of the largest investigations of 1,108 unique children with DS based at a single institution incorporating a multi-age and ethnic sample and current AAP Guidelines testing practices

In conclusion, this study reinforces the unique phenotypic clinical pattern seen in individuals with DS and the importance of utilizing and expanding the AAP’s DS Guidelines in the care for children and young adults with DS. This information can aid in determining appropriate referrals that improve outcomes for our patients and provide up to date information on DS to families and providers. Data from the SCDS Clinic Database led to the adjustment of care, including initiating the start of three unique clinics: Infant, Feeding and Sleep. Clinic outreach efforts in the community may be needed to target underserved groups, such as the Hispanic and Latino population and develop relationships for medical needs. Pediatricians and family practice physicians need to be aware of: (1) The increased incidence of comorbidities, especially at birth, (2) the importance of appropriate evaluation and referral consistent with the compliance to the AAP’s DS Guidelines for all children with DS, (3) establishing treatment plans with a developmental specialty team for identified comorbidities and (4) implementing routine celiac screenings and swallow studies not currently outlined in the AAP DS Guidelines to meet the medical needs of children with DS.

We would like to thank the Global Down Syndrome Foundation and the Anna and John J. Sie Foundation for their financial support of research conducted at the Sie Centre for Down Syndrome at Children’s Hospital Colorado.