Prevalence of Down Syndrome Among Children and Adolescents in 10 Regions of the United States
OBJECTIVE: We aimed to estimate the prevalence of Down syndrome (DS) among children and adolescents aged 0 to 19 years in 10 regions of the United States.
METHODS: This study was a cross-sectional analysis of live-born infants with DS during 1979–2003 from 10 population-based birth defects registries in the United States. We estimated the prevalence of DS at birth and among children aged 0 to 19 years in each region and in all regions pooled. The prevalence of DS among children and adolescents was calculated overall and according to age group, race/ethnicity, infant gender, and presence of a major heart defect.
RESULTS: From 1979 through 2003, the prevalence of DS at birth increased by 31.1%, from 9.0 to 11.8 per 10000 live births in 10 US regions. In 2002, the prevalence among children and adolescents (0–19 years old) was 10.3 per 10000. The prevalence of DS among children in a given age group consistently increased over time but decreased with age within a given birth cohort. The pooled prevalence of DS among children and adolescents was lower among non-Hispanic black individuals and other racial/ethnic groups compared with non-Hispanic white individuals; it was also lower among females than males.
CONCLUSIONS: This study provides prevalence estimates of DS among children and adolescents from 10 US regions. These estimates varied according to region, race/ethnicity, and gender, suggesting possible variation in prevalence at birth or in survival rates on the basis of these characteristics.
Down syndrome (DS) is the most common chromosomal disorder with an estimated 5400 infants with DS born each year in the United States.1,2 For children with DS, the chance of survival beyond the first year of life has improved in recent years,2–7 with 90% of children with DS now surviving beyond 5 years of age.6 Children with DS have an increased risk for endocrinologic, hematologic, respiratory, and neurologic sequelae, as well as psychiatric and social problems later in life.8,9 Given these potential problems and that transitioning from pediatric to adult health care services is challenging for children with disabilities,10–12 planning for appropriate health and other services is likely to require information on the number of children who have DS and are living in a given community. To date, however, only 1 study in a metropolitan area has estimated the prevalence of DS among children and adolescents, reporting a prevalence of 8.3 per 10000 children aged 0 to 19 years.13
In this article, we estimate DS prevalence among children and adolescents in 10 regions of the United States. We also report findings on the variation in DS prevalence according to region, age group, and factors that have been reported to be associated with variations in DS prevalence among children.
Data from 10 population-based birth defects surveillance programs were used to estimate DS prevalence among children and adolescents aged 0 to 19 years by using methods previously described for a study conducted in Atlanta.13 Data on the number of live-born infants with DS were provided from the following regions and birth years: Arkansas (1993–2002), Georgia (5 central counties of metropolitan Atlanta, 1979–2003), California (11 counties, 1983–2002), Colorado (1989–2003), Iowa (1983–2003), North Carolina (1989–1993 and 1995–2003), New York (New York City excluded, 1983–2003), Oklahoma (1994–2003), Texas (1994–2003), and Utah (1995–2003). Nine of these programs collected birth defects data by using active or a combination of active and passive case-finding methods, whereas Colorado used passive case-finding methods (vital records or hospital data). Seven regions used 6-digit codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) or modified codes from the British Paediatric Association Classification of Diseases (BPA) to ascertain cases. North Carolina (1989–2003), Colorado (1989–2003), and New York (1983–1991) used 4- or 5-digit codes from the previous ICD-9. Details about these surveillance and coding systems have been published elsewhere.14,15
In our study, we included live-born infants with DS when they were ≥20 weeks of gestational age or had a birth weight of ≥500 g and had a DS diagnosis indicated by ICD-9-CM or BPA codes. We included only cytogenetically confirmed cases from 7 regions (Arkansas, Georgia, California, Iowa, New York, Oklahoma, and Utah) and all cases (cytogenetically confirmed and unconfirmed) from 3 regions (Colorado, North Carolina, and New York [1983–1991]) that did not indicate cytogenetic confirmation in their coding systems. Infants with DS were classified as having congenital heart defects (CHDs) when their records indicated the presence of ICD-9-CM or BPA codes for heart defects (745.000–747.430 and 747.640). For infants who had DS and had CHD codes that lacked specificity or included both major and minor cardiac defects, a pediatric cardiologist helped us to classify and decide whether to include these cases in the CHD group. For infants who had DS and were included in the study, information on gender (female, male) and maternal race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, and other) was obtained from birth defect registries, and vital status through June 30, 2002, was determined from hospital records, state vital records, and a probability-based linkage with the National Death Index.13,16 Because information on migration out of the region for children with DS was not available, we assumed that children with DS resided in their birth region through 2002 or that for a given number of children who had DS and migrated out of a given region, a similar number of children with DS and similar characteristics moved into the region (ie, 0 net migration).
For each study year (1979–2003), DS prevalence per 10000 live births was calculated for each region and for all regions pooled as the sum of the number of live-born infants with DS divided by the sum of the number of live-born infants in the underlying population, as ascertained by each registry. We used Poisson regression and the pooled prevalence according to year to calculate the average annual percentage change (AAPC) in DS prevalence at birth overall and stratified according to maternal age (<35 or ≥35 years). To examine variations in DS prevalence at birth, we calculated overall 5-year birth prevalence (1999–2003) according to region and further stratified regional prevalence according to race/ethnicity, infant gender, presence of CHDs, and maternal age. We used Poisson regression to calculate prevalence ratios (PRs) and 95% confidence intervals (CIs) by using SAS 9.1 (SAS Institute Inc, Cary, NC).
We estimated DS prevalence among children in various age groups in 2002 on the basis of the available birth cohorts in each region and for all regions pooled by using the same methods applied in a previous study.13 For each region, the numerator was the children who were born between July 1, 1982, and June 30, 2002, and alive as of July 2002, and the denominator was the total number of children in the US Census population for the corresponding region in July 2002.13 Because long-term ascertainment of births with DS was limited for most registries except Georgia, estimates of DS prevalence for the oldest age group (16–19 years) could be calculated only for Georgia. We also estimated prevalence among children who had DS and were aged 0 to 19 years during 1990–2002 to examine temporal trends according to age group.
Using Poisson regression, we estimated adjusted PRs for age group, race/ethnicity, and infant gender within each region and then estimated the adjusted PRs across 10 regions by using a meta-likelihood method,17,18 which accounts for regional variations in PRs. We did not adjust for presence of CHDs because the quality of the CHD data is questionable for some registries as a result of less specific diagnostic coding systems.
From 1979 to 2003, the pooled DS prevalence at birth across the 10 US regions increased from 9.0 to 11.8 per 10000 live births, representing an average increase of 0.9% per year (P < 0.001) (Fig 1). During the same period, prevalence at birth significantly increased among births to older mothers (AAPC: 1.9%; P < .0001) and decreased slightly among births to younger mothers (AAPC: −0.6%; P = .0007).
For the most recent 5-year period (1999–2003), the pooled DS prevalence at birth was 11.8 per 10000 (Table 1), with prevalence varying according to region and ranging from 9.7 in Arkansas to 13.7 in Utah. The overall DS prevalence at birth was almost 5 times higher among births to older mothers (38.6 per 10 000) than among births to younger mothers (7.8 per 10 000). The pooled DS prevalence at birth also varied significantly according to race/ethnicity, infant gender, and presence of CHDs; these variations also were evident in some regions. DS prevalence at birth was consistently higher among non-Hispanic white than among non-Hispanic black individuals, but this disparity reached statistical significance only in Texas. In 4 regions (Arkansas, California, North Carolina, and Texas), DS prevalence at birth was higher among Hispanic than among non-Hispanic white individuals. Only 3 regions (California [n = 115], Oklahoma [n = 65], and Texas [n = 312]) had >30 children and adolescents of other race/ethnicity, limiting meaningful analysis of regional variation for this group. The DS prevalence at birth was significantly higher among males than among females in California, Colorado, and Texas. Overall, there was a higher prevalence of infants with DS and CHDs than of those without CHDs, but this finding varied according to region.
In 2002, the pooled DS prevalence among children and adolescents aged 0 to 19 years was 10.3 per 10 000 (Table 2). DS prevalence varied according to region when stratified according to age group, race/ethnicity, infant gender, and presence of CHDs. The pooled DS prevalence declined from the youngest to the oldest age groups and was significantly lower among non-Hispanic black individuals (PR: 0.74 [95% CI: 0.66–0.82]) and among children and adolescents of other race/ethnicity (PR: 0.84 [95% CI: 0.71–0.99]) when compared with non-Hispanic white individuals. DS prevalence was lower among children and adolescents with CHDs than among those without CHDs (crude PR: 0.83 [95% CI: 0.80–0.86]). The pooled prevalence among Hispanic individuals was no longer significantly different from non-Hispanic white individuals after adjustment for age group and gender. The adjusted pooled prevalence was 9% lower among females than among males.
In 2002, DS prevalence declined with increasing age in all racial/ethnic groups, but the decline was more apparent among Hispanic (43%) than among non-Hispanic white individuals (19%; Fig 2). DS prevalence was consistently higher among males than among females regardless of race/ethnicity or age group.
In 2002, DS was present in 1 of every 971 children and adolescents who were aged 0 to 19 and living in 10 US regions. DS prevalence at birth and among children and adolescents increased significantly from 1990 to 2003 but decreased with age within each birth cohort. DS prevalence among children and adolescents varied according to region, age group, race/ethnicity, and infant gender.
Our estimated DS prevalence at birth (11.8 per 10 000 live births during 1999–2003) included only live-born infants with DS in the numerator, whereas the national estimates by Canfield et al1 included DS among fetal deaths (12.9 per 10 000 live births reported from 11 active surveillance systems during 1999–2001). DS prevalence at birth was significantly lower among non-Hispanic black individuals and was higher among Hispanic individuals compared with the prevalence among non-Hispanic white individuals, which is consistent with previous findings.1,13,19,20 Our study also confirmed an increase in prevalence in DS at birth over time, and this trend over time paralleled the increasing proportion of births to older mothers (aged ≥35 years) in the 10 US regions.
We also found an increasing prevalence of DS among children and adolescents, with a higher prevalence evident among younger than among older age groups. These findings could reflect 1 or several factors, including an increasing prevalence of DS among live births, an increasing proportion of births to older mothers, and recent improvements in the survival of infants with DS related to earlier surgical interventions for CHD.5–8,13,21 Our prevalence estimate beyond infancy suggests that the number of children and adolescents who were aged 0 to 19 years and living with DS in the United States in 2002 was ∼83 400 (95% CI: 82 846–83 980).
Among race/ethnic groups, we observed the lowest prevalence of DS among non-Hispanic black individuals aged 0 to 19 years, which may reflect a lower prevalence of DS at birth, underdiagnosis of DS prenatally or at birth, differences in access to treatment, and/or lower survival rates among non-Hispanic black individuals.4,7,13,22–25 Although the pooled DS prevalence among children and adolescents seemed to be higher among Hispanic than among non-Hispanic white individuals, this variation disappeared after controlling for age and gender. In stratified analysis, prevalence estimates were similar in the youngest age group for Hispanic and non-Hispanic white individuals; however, the prevalence of DS declined more rapidly with increasing age among Hispanic than among non-Hispanic white individuals in our analysis. Possible reasons for this trend include a higher DS prevalence at birth in recent birth cohorts and/or lower survival rates among Hispanic individuals with DS in earlier birth cohorts. Because Hispanic immigrant women tend to have higher rates of fertility, late entry to prenatal care, and lower level of insurance coverage,26 it is important to conduct more research to elucidate the extent to which these factors might be associated with the observed trends in prevalence and survival in states with rapid Hispanic population growth.
One strength of our study is the large sample size of infants born with DS provided by the 10 population-based birth defects surveillance systems, which collected data on ∼29% of all births in the United States in 2002.27 Data from 10 regions allowed us to examine ethnic diversity in the prevalence of DS including Hispanic, as well as to estimate the DS prevalence according to age group and region. Our race/ethnic-, age-, and gender-specific prevalence may be useful to local health care providers to plan the resources needed to treat each age group effectively and to project future needs as the population of children with DS ages and the prevalence of adults with DS increases.
Our study had some limitations. The number of children living with DS in 2002 was estimated from determinations of vital status rather than on actual tracing of the living children; however, we obtained vital status information from 3 sources (medical records, state vital records, and the National Death Index), which probably helped to minimize potential errors in undercounting deaths. Census data are the best available denominator data for our prevalence estimates13 because they account for birth, death, and migration at the county level; however, according to US Census data, >16% of residential moves in the United States occur among children aged <20 years.28 For the numerator of our prevalence estimates, we had no information on residential mobility; therefore, we had to assume a net migration of 0 for children with DS for each study region.13 If the net migration of children with DS into the 10 regions happened to be positive, then our estimates of prevalence would be underestimates. If the net migration happened to be negative, then the converse would be true.
In addition, although we used meta-analysis to minimize regional variation, the heterogeneity of data presented some challenges for estimating the pooled prevalence among children with DS. For example, CHDs among children with DS in our study ranged from 37% (Georgia) to 61% (Oklahoma), which probably reflects differences in ascertainment methods, coding system, and data quality across birth defect registries, rather than sampling variations.29 Our estimates in the oldest age group (12–19 years) were limited to surveillance programs that have been in operation for ≥12 years and for periods of consistent data collection, which meant restricting analyses to a smaller number of regions and time periods. Continued state and federal investment in, refinement of, and improved standardization across in the regional birth defects surveillance programs would allow for more precise estimates in the future. Until then, our methods are the best compromise to date for generating national estimates of DS among children and adolescents.
Because DS prevalence at birth is increasing and the median age at death for individuals with DS has increased in recent years (from 25 years in 1983 to 49 in 1997) in the United Sates,4 the number of children and adolescents who have DS and live in the United States is likely to increase over the next several years. As an increasing number of children and adolescents with DS age, a number of health care issues related to transition to adulthood (eg, how to ensure continuity of care for conditions such as complex congenital heart defects) and quality of life are likely to gain increasing prominence. Efforts to collect population-based data on DS prevalence and longitudinal data on the occurrence of secondary conditions and determinants of such conditions among children and adolescents with DS will represent an initial step in the development of evidence-based strategies to address such issues.13,30–32
Our study represents one of the first efforts to provide population-based estimates of DS prevalence among children and adolescents in the United States. Determination of whether health services are meeting the needs of an increasing number of aging individuals with DS to ensure a healthy life across the life span is likely to require more extensive data linkages of health databases on individuals with DS and considerations of the development and maintenance of long-term longitudinal DS registries. Our population-based estimates of the prevalence of DS among children and adolescents in 10 US regions could be useful for estimating the number of affected children and adolescents at the community and state levels.
We thank the National Birth Defects Prevention Network and, in particular, the staff of the following birth defects monitoring programs, which participated in the Congenital Anomaly Multistate Prevalence and Survival (CAMPS) Collaborative for their conscientious and skilled data-collection efforts and contributions that made this study possible: Arkansas Reproductive Health Monitoring System (Charlotte Hobbs and Bridget S. Mosley), California Birth Defects Monitoring Program (Barbara Warmerdam, Marcia Ehinger, and Gary Shaw), Colorado Responds to Children With Special Needs (Carol Stanton and Margaret Ruttenber), Iowa Registry for Congenital and Inherited Disorders (Bradley D. MacDowell and Paul Romitti), Metropolitan Atlanta Congenital Defects Program, New York State Congenital Malformations Registry (Charlotte Druschel and Ying Wang), North Carolina Birth Defects Monitoring Program (Katie Harmsen and Robert Meyer), Oklahoma Birth Defects Registry (Kay A. Pearson), Texas Birth Defects Epidemiology and Surveillance Branch (Lisa K Marengo and Mark Canfield), and Utah Birth Defect Network (Marcia Feldkamp and Miland Palmer). We thank Dr Tiffany Riehle-Colarusso, Mike Atkinson, and Don Gambrell for expertise and technical assistance in case classification and data linkages, respectively, and all Metropolitan Atlanta Congenital Defects Program staff for constructive suggestions.
- Accepted July 9, 2009.
- Address correspondence to Mikyong Shin, DrPH, MPH, RN, Centers for Disease Control and Prevention, 1600 Clifton Rd, Mail Stop E-86, Atlanta, GA 30333. E-mail:
This study was presented at the 11th annual meeting of the National Birth Defects Prevention Network; February 11–13, 2008; Washington, DC.
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Financial Disclosure: The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject:
There has been only 1 published study on estimates of the prevalence of DS among children and adolescents in 1 metropolitan area of the United States.
What This Study Adds:
This is the first study to provide population-based estimates of prevalence of DS among children and adolescents aged 0 to 19 years for 10 diverse regions of the United States.
- ↵Bittles AH, Bower C, Hussain R, Glasson EJ. The four ages of Down syndrome. Eur J Public Health.2007;17 (2):221– 225
- ↵Halliday J, Collins V, Riley M, Youssef D, Muggli E. Has prenatal screening influenced the prevalence of comorbidities associated with Down syndrome and subsequent survival rates? Pediatrics.2009;123 (1):256– 261
- ↵National Birth Defects Prevention Network. State birth defects surveillance program directory. Birth Defects Res Part A Clin Mol Teratol.2008;82 (12):907– 961
- ↵Centers for Disease Control and Prevention. National Death Index User Manual. Hyattsville, MD: US Department of Health and Human Services, National Center for Health Statistics; 2000
- ↵SAS Institute Inc. SAS/IML Softwares: Changes and Enhancements [computer software]. Release 8.2. Cary, NC: SAS Institute Inc; 2001
- ↵Bishop J, Huether CA, Torfs C, Lorey F, Deddens J. Epidemiologic study of Down syndrome in a racially diverse California population, 1989–1991. Am J Epidemiol.1997;145 (2):134– 147
- ↵US Census Bureau. Geographic morbidity: 2002 to 2003 (P20-549). Available at: www.census.gov/population/www.socdemo/migrate.html. Accessed July 2008
- Henderson A, Lynch SA, Wilkinson S, et al. Adults with Down's syndrome: the prevalence of complications and health care in the community. Br J Gen Pract.2007;57 (534):50– 55
- ↵Rasmussen SA, Whitehead N, Collier SA, Frias JL. Setting a public health research agenda for Down syndrome: summary of a meeting sponsored by the Centers for Disease Control and Prevention and the National Down Syndrome Society. Am J Med Genet A.146A (23):2998– 3010, 2008
- Copyright © 2009 by the American Academy of Pediatrics