National time trends in congenital heart defects, Denmark, 1977-2005
Article Outline
- Abstract
- Methods
- Results
- Discussion
- Disclosures
- Acknowledgements
- Appendix A. Definition of cardiac phenotypes
- References
- Copyright
Background
Time trends in congenital heart defects (CHD) by specific phenotype and with long follow-up time are rarely available for an entire population. We present trends in national CHD prevalences over the past 3 decades.
Methods
We linked information from the National Patient Register, the Causes of Death Register, and the Danish Cytogenetic Central Register for all persons born in Denmark, 1977 to 2005, and registered in the Civil Registration System, yielding a cohort of 1,763,591 persons—18,207 with CHD. Individuals with CHDs were classified by phenotype (heterotaxia, conotruncal defect, atrioventricular septal defect, anomalous pulmonary venous return, left and right ventricular outflow tract obstructions, septal defects, complex defects, associations, patent ductus arteriosus, unspecified, and other specified) by combining International Classification of Diseases codes using a hierarchical approach.
Results
From 1977 to 2005, the overall CHD birth prevalence increased from 73 to 113 per 10,000 live births. Generally, prevalence increased for defects diagnosed in infancy, until 1996–1997, and then stabilized. For each 5-year interval, isolated septal defects and severe defects increased by 22% (95% CI, 20%-25%) and 5% (95% CI, 4%-7%), respectively. Among the severe defects, conotruncal defects and atrioventricular septal defect showed the largest prevalence increases. Women had a lower prevalence of severe defects during the 1980s. The CHD prevalence increase was unchanged when persons with extracardiac defects or chromosomal aberrations were excluded.
Conclusions
CHD birth prevalence increased from the beginning of the 1980s but stabilized in the late 1990s.
Congenital heart defects (CHD), defined as gross structural abnormalities of the heart or intrathoracic vessels that are actually or potentially of functional significance,1 are among the most common birth defects, affecting 5 to 10 per 1,000 live births.2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 The birth prevalence has increased since the 1980s, likely due to better case ascertainment or improved diagnostics.5, 14 The CHD prevalence in the total population has also increased because of the improved management of affected persons.9, 15 At present, there are similar numbers of adults and children with severe malformations9, 10; the population CHD prevalence is 4 per 1,000 persons overall and 0.4 per 1,000 persons for severe CHDs. With the birth prevalence and the population prevalence to the US population,16 20,000 to 40,000 new CHD cases arise per year, and there are 1,200,000 prevalent cases of CHD at any point in time, of which 200,000 have a severe heart malformation. The etiology of CHD is largely unknown,8, 17 and the burden imposed on affected persons, their families, and society at large is considerable.
Epidemiologic studies of CHDs have not been undertaken at the national level in Denmark since 1980,18 although large population-based health registers are available. One reason could be that the classification of CHDs is challenging because of the wide range of known defects, with different underlying etiologies for each class of defects. In surveillance registers, such as the European Surveillance of Congenital Anomalies,19 and medical birth registers, individuals with multiple cardiac defects are usually counted several times. The alternative, assigning each individual (including those with multiple cardiac defects) to a specific cardiac phenotype,20 is more relevant to etiologic research but is also more difficult.
At the National Birth Defects Prevention Group in Atlanta, GA, Botto et al20 have developed a classification scheme of cardiac defects according to heart anatomy and fetal heart development stages in a flexible manner allowing CHDs to be combined in different ways according to new hypotheses. Inspired by the Botto et al classification of CHDs, we developed an algorithm for classification of CHDs registered in Denmark's National Patient Register. By combining International Classification of Diseases (ICD) codes in a hierarchical fashion, individuals with CHDs were classified into heart phenotypes.
The purposes of the present study were to classify all registered CHDs in Denmark in the period 1977 to 2005; to evaluate birth prevalence of CHDs by calendar period, age at diagnosis, sex, and type of defect; and to establish a database for further research.
Methods
Data sources
Since April 1, 1968, the Danish Civil Registration System has registered demographic, residence, vital status, and kinship information on all persons residing in Denmark, aided by the unique personal identification number assigned to each Danish resident. The personal identification number permits accurate linkage of individual-level information from Denmark's nationwide population-based registers, including the National Patient Register (NPR), the Medical Birth Register, the Causes of Death Register (CDR), and the Danish Cytogenetic Central Register (DCCR).
The NPR contains information on in-patient diagnoses and operations performed since 1977 and outpatient diagnoses from 1995 onward. The Medical Birth Register has collected information on gestational age for all births since 1978. The CDR contains death certificate information including underlying cause of death and up to 3 contributing causes of death for the period 1970 to 2004. The DCCR was established in 1968 and contains information on all pre- and postnatal chromosomal analyses performed in Denmark since 1970 and 1960, respectively.
Case ascertainment and classification of CHDs
Congenital heart defects were ascertained using the NPR and the CDR, using ICD, Eight Revision (ICD-8) codes for diagnoses registered from 1977 through 1993 (746.00–747.49, 759.00, 759.01, 759.09) and ICD, Tenth Revision (ICD-10) codes (Q200-Q269, Q893) thereafter. We considered an individual to have a CHD if a defect was ever diagnosed, irrespective of age at diagnosis. Stillbirths were not included in the present study because no cardiac information was available.
Congenital heart defects were classified into phenotypes based on those used by Botto et al.20 Persons with CHDs were classified into the following 17 heart phenotypes by grouping specific ICD codes in hierarchical fashion: (1) heterotaxia; (2) conotruncal defect; (3) atrioventricular septal defect (AVSD); (4) anomalous pulmonary venous return (APVR); (5) left ventricular outflow tract obstruction (LVOTO); (6) right ventricular outflow tract obstruction (RVOTO); (7) isolated atrial septal defect (ASD); (8) isolated ventricular septal defect (VSD); (9) ASD and VSD; (10) complex defects; (11) conotruncal defect + AVSD; (12) septal defect + LVOTO; (13) septal defect + RVOTO; (14) isolated patent ductus arteriosus (PDA) in infants born at term; (15) isolated PDA in preterm infants; (16) unspecified; and (17) all other specified CHDs. Details on the CHD classification are placed in the Appendix (available online).
In addition, persons with CHDs were categorized as having septal defects only, severe defects (heterotaxia, conotruncal defects, AVSD, APVR, LVOTO, RVOTO, complex defects, associations (phenotypes 11-13), other specified CHDs, or unspecified CHDs. Persons with CHD were also divided into those with and without extracardiac defects (ICD-8: 740-759; ICD-10: Q00-Q99 other than CHD codes), and those with and without chromosomal aberrations.
Chromosomal aberrations
Information on reported chromosomal aberrations (Down syndrome, trisomy 13, trisomy 18, Turner syndrome, other sex chromosome aneuploidy, deletions, and other chromosome abnormalities) was collected from DCCR.
Study population
All individuals born in Denmark in 1977 to 2005 were included in the study cohort. Date of birth for all cohort members was identified using the Civil Registration System, and information on CHDs was retrieved from the NPR and the CDR.
Statistical analysis
Absolute risk of CHD was estimated in the entire population using the prevalence among live births, that is, the number of live births with CHD divided by the total number of live births (birth prevalence), and reported with a 95% CI. Time trends for CHDs were evaluated in singletons. Defect prevalences by calendar period (1977-79, 1980-1984, 1985-1989, 1990-1994, 1995-1999, and 2000-2005) were tested for trend by assigning values 1 through 6 to the 6 categories of calendar period and analyzing calendar period as a continuous variable in a log-linear binomial regression model. Congenital heart defect prevalences were stratified by age at diagnosis and sex. All analyses were performed in SAS (9.1, SAS Institute Inc, Cary, NC).
The study was funded by Lundbeck Foundation R17-A1479, Copenhagen, Denmark and by the Western Norway Regional Health Authority 911280, Bergen, Norway.
Results
Of 1,763,591 persons born in Denmark in the period 1977 to 2005, 18,207 had ≥1 CHDs, yielding an overall CHD prevalence of 103 per 10,000 live births (Table I). Isolated septal defects were the most prevalent defects, comprising nearly one third of all CHDs. Conotruncal defect prevalence was 11.5 per 10,000 live births. Atrioventricular septal defect had a prevalence of 4.46 per 10,000 live births, and the left and right heart obstructive defects had prevalences of 6.35 and 3.61 per 10,000 live births, respectively. Heterotaxia had a prevalence of 1.78 per 10,000 live births. Unspecified CHDs represented one seventh of the CHDs. Among persons with a CHD, 1,430 (7.9%) were part of a multiple birth, with the highest proportion of multiple births observed among persons with isolated PDA born preterm. Additional extracardiac birth defects were found in 4,067 (22.3%) persons with CHD. Among persons with CHDs, 3,097 (17.0%) had a postnatal chromosomal investigation and of these 1,272 (41.1%) had a chromosomal aberration. Persons with AVSD were most likely to have such aberration (Table I).
Table I. Congenital heart defects in 1 763 591 persons born in Denmark—1977 to 2005†
| CHD | Total number of persons | Multiple birth | Extracardiac defect | Chromosomal aberration | Prevalence per 10 000 live births |
|---|---|---|---|---|---|
| No. (%) | No. (%) | No. (%) | |||
| Any CHD | 18 207 | 1430 (7.9) | 4067 (22.3) | 1,272 (7.0) | 103.2 |
| Heterotaxia | 314 | 10 (3.2) | 143 (45.5) | 60 (19.1) | 1.78 |
| Conotruncal defect | 2031 | 90 (4.4) | 578 (28.5) | 160 (7.9) | 11.5 |
| AVSD | 787 | 33 (4.2) | 208 (26.4) | 331 (42.1) | 4.46 |
| APVR | 160 | 9 (5.6) | 47 (29.4) | 4 (2.5) | 0.91 |
| LVOTO | 1119 | 67 (6.0) | 206 (18.4) | 46 (4.1) | 6.35 |
| RVOTO | 637 | 36 (5.7) | 143 (22.4) | 17 (2.7) | 3.61 |
| Septal defect, isolated | 6073 | 401 (6.6) | 1282 (21.1) | 389 (6.4) | 34.4 |
 ASD | 2178 | 199 (9.1) | 536 (24.6) | 172 (7.9) | 12.3 |
| VSD | 3295 | 159 (4.8) | 579 (17.6) | 137 (4.2) | 18.7 |
| ASD + VSD | 600 | 43 (7.2) | 167 (27.8) | 80 (13.3) | 3.40 |
| Complex defect | 40 | 2 (5.0) | 15 (37.5) | 4 (10.0) | 0.23 |
| Associations | 348 | 28 (8.0) | 82 (23.6) | 25 (7.2) | 1.97 |
| Conotruncual + AVSD | 32 | 2 (6.3) | 15 (46.9) | 14 (43.8) | 0.18 |
| Septal + LVOTO | 166 | 9 (5.4) | 39 (23.5) | 4 (2.4) | 0.94 |
| Septal + RVOTO | 150 | 17 (11.3) | 28 (18.7) | 7 (4.7) | 0.85 |
| PDA, isolated | 2,803 | 555 (19.8) | 583 (20.8) | 61 (2.2) | 15.9 |
| At term | 759 | 28 (3.7) | 173 (22.8) | 37 (4.9) | 4.3 |
| Preterm | 1762 | 480 (27.2) | 348 (19.8) | 10 (0.6) | 10.0 |
| Gestational age unknown⁎ | 282 | 47 (16.7) | 62 (22.0) | 14 (5.0) | 1.60 |
| Unspecified CHD | 2564 | 117 (4.6) | 447 (17.4) | 111 (4.3) | 14.5 |
| Other CHD | 1331 | 82 (6.2) | 333 (25.0) | 64 (4.8) | 7.55 |
⁎Gestational age unavailable for births in 1977. |
†Among persons with CHDs, 3,097 (17.0%) had a postnatal chromosomal investigation and of these 1,272 (41.1%) had a chromosomal abberation. |
Overall, CHD prevalences increased over the observation period (Table II), with CHD prevalence increasing by 10% per 5-year interval. Among singletons, the prevalence increase was smaller (8% per 5-year interval). The percent change in CHD prevalences for males and females were 8% and 9%, respectively.
Table II. Prevalence⁎ of congenital heart defects in 1 763 591 persons born in Denmark—1977 to 2005
| Prevalence⁎ (95% CI) | % Change in prevalence per 5-y interval (95% CI) | ||||||
|---|---|---|---|---|---|---|---|
| 1977-1979 | 1980-1984 | 1985-1989 | 1990-1994 | 1995-1999 | 2000-2005 | ||
| n = 183 769 | n = 266 367 | n = 286 275 | n = 334 026 | n = 338 846 | n = 354 308 | ||
| All individuals | 73.3(69.5-77.3) | 85.4 (82.0-89.0) | 90.1 (86.7-93.6) | 115 (111-119) | 123 (119-127) | 113 (110- 117) | 10% (9-11) |
| Singletons† | 71.7 (67.9-75.7) | 82.5 (79.1-86.1) | 86.9 (83.5-90.4) | 110 (106-113) | 117 (113-120) | 104 (101- 108) | 8% (7-9) |
| Men | 73.0 (67.7-78.7) | 83.8 (79.0-88.8) | 91.9 (87.1-97.0) | 108 (103-113) | 117 (112-122) | 103 (98.7-108) | 8% (6-9) |
| Women | 70.4 (65.1-76.1) | 81.2 (76.4-86.3) | 81.5 (76.9-86.4) | 111 (106-117) | 117 (111-122) | 105 (100-110) | 9% (8-11) |
| Excluding extracardiac defects | 56.2 (52.8-59.7) | 61.6 (58.7-64.7) | 65.4 (62.4-68.4) | 85.5 (82.4-88.7) | 91.0 (87.8-94.3) | 82.4 (79.4-85.5) | 9% (8-11) |
| Excluding chromosomal anomalies | 66.3 (62.6-70.1) | 77.2 (73.9-80.6) | 80.2 (77.0-83.6) | 102 (98.9-106) | 108 ( 105-112) | 95.3 (92.1-98.7) | 8% (7-9) |
⁎Prevalence per 10,000 live births. |
In Table III, prevalences of the different heart phenotypes in singletons are presented by calendar period. For the isolated septal defects, the change in prevalence was 22% per 5-year interval, that is, prevalence more than doubled over the 29-year study period. For each specific septal defect (ASD, VSD, and ASD + VSD), the changes in prevalence were 30%, 15%, and 50%, respectively. Septal defects in combination with RVOTO also increased in prevalence, whereas septal defects in combination with LVOTO did not. The prevalence of other severe defects changed little over the study period. Changes in heterotaxia, APVR, and LVOTO prevalences over the study period were negligible, whereas the prevalences of conotruncal defects, AVSD, and RVOTO increased slightly (4% to 9% per 5-year interval). The prevalence of complex defects increased by 40% per 5-year interval, but only 38 such defects were diagnosed among singletons during the entire study period. Likewise, only 30 singletons had conotruncal defects in combination with AVSD. For all severe defects combined, the overall prevalence increased by 5% (95% CI, 4%-7%) per 5-year interval. Isolated PDA prevalence increased by 9% per 5-year interval. During the 1990s, there was a major shift in the prevalences of other specified CHDs and unspecified defects, with the former stabilizing after having increased in earlier decades and the latter decreasing dramatically after previously having been stable.
Table III. Prevalence⁎ of CHD by birth cohort in 1,711,641 singletons born in Denmark—1977 to 2005
| CHD | Prevalence | % Change in prevalence per 5-y interval (95% CI) | Prevalence | |||||
|---|---|---|---|---|---|---|---|---|
| 1977-1979 | 1980-1984 | 1985-1989 | 1990-1994 | 1995-1999 | 2000-2005 | 1977-2005 | ||
| n = 180 155 | n = 260 989 | n = 280 016 | n = 324 501 | n = 326 514 | n = 339 466 | N = 1 711 641 | ||
| All CHDs | 71.7 | 82.5 | 86.9 | 110 | 117 | 104 | 8% (7-9) | 98.0 |
| Heterotaxia | 1.33 | 1.72 | 1.68 | 1.82 | 2.08 | 1.80 | 5% (−2 to 13) | 1.78 |
| Conotruncal defect | 9.55 | 10.5 | 11.4 | 11.8 | 11.9 | 11.8 | 4% (1-7) | 11.3 |
| AVSD | 4.11 | 3.37 | 3.96 | 4.41 | 5.15 | 5.01 | 8% (3-13) | 4.41 |
| APVR | 0.67 | 0.88 | 0.86 | 0.86 | 0.80 | 1.12 | 7% (−3 to 18) | 0.88 |
| LVOTO | 6.38 | 6.32 | 5.86 | 5.67 | 6.40 | 6.33 | 0% (−3 to 4) | 6.15 |
| RVOTO | 2.16 | 2.53 | 3.71 | 4.31 | 4.10 | 3.48 | 9% (4-15) | 3.51 |
| Septal defect isolated | 18.7 | 20.9 | 22.6 | 35.3 | 46.3 | 44.2 | 22% (20-25) | 33.1 |
| ASD | 6.16 | 6.48 | 6.71 | 11.7 | 15.9 | 18.0 | 30% (26-34) | 11.6 |
| VSD | 11.8 | 13.0 | 14.5 | 20.5 | 24.8 | 20.7 | 15% (12-17) | 18.3 |
| ASD + VSD | 0.67 | 1.38 | 1.36 | 3.11 | 5.60 | 5.51 | 50% (41-59) | 3.25 |
| Complex | 0.06 | 0.11 | 0 | 0.37 | 0.37 | 0.29 | 40% (12-75) | 0.22 |
| Associations | 1.11 | 1.53 | 0.93 | 2.19 | 2.42 | 2.47 | 19% (11-28) | 1.87 |
| Conotruncal + AVSD | 0.22 | 0.15 | 0.21 | 0.12 | 0.21 | 0.15 | −4% (−23 to 20) | 0.18 |
| Septal + LVOTO | 0.89 | 1.11 | 0.43 | 0.89 | 1.04 | 1.09 | 5% (−5 to 16) | 0.92 |
| Septal + RVOTO | 0 | 0.27 | 0.29 | 1.17 | 1.16 | 1.24 | 51% (33-71) | 0.78 |
| PDA, isolated | 5.83 | 12.5 | 13.0 | 15.7 | 14.7 | 13.6 | 9% (6-12) | 13.1 |
| At term | 1.50 | 3.83 | 3.96 | 4.50 | 5.42 | 5.01 | 15% (10-21 | 4.27 |
| Preterm | 1.50 | 6.82 | 7.93 | 9.46 | 8.94 | 7.54 | 13% (9-17) | 7.49 |
| Unknown | 2.83 | 1.88 | 1.14 | 1.79 | 0.34 | 1.00 | −22% (−28 to –16) | 1.37 |
| Unspecified CHDs | 17.2 | 17.2 | 17.7 | 18.1 | 12.9 | 5.33 | –15% (–17 to –13) | 14.3 |
| Other specified CHDs | 4.66 | 4.90 | 5.07 | 9.00 | 9.43 | 8.69 | 17% (13-21) | 7.30 |
⁎Prevalence per 10,000 live singleton births. |
As shown in Figure 1, the overall prevalence of CHDs among singletons increased from 1977-79, peaked in the period 1996–1997, and then declined slightly. Stratifying CHD prevalence by age at diagnosis showed that increases in numbers of defects diagnosed in infancy contributed to the prevalence increase up to 1996–1997. Decreases in numbers of defects diagnosed after age five years and at 1 to 4 years of age contributed to the observed declines in prevalence after 1996–1997 and 2000, respectively; these decreases were likely at least partly due to shorter follow-up time in children born at the end of the study period. After 1996–1997, the prevalence of defects diagnosed in infancy stabilized. Sex-specific prevalence curves had overlapping 95% CIs (data not shown).
Figure 1.
Prevalence of congenital heart defects by birth cohort and age at diagnosis in 1,711,641 singletons born in Denmark—1977 to 2005.
Figure 2 shows the prevalences of all CHDs, isolated septal defects, severe defects, other specified defects, and unspecified defects. In the late 1980s, the prevalence of both septal defects and severe defects began increasing (at a rate of 22% and 5% per 5-year interval, respectively). Most of the increases in septal and severe defect prevalences were due to increases in the numbers of such defects diagnosed in infants (data not shown). From the beginning of the 1990s, the prevalence of other specified defects increased by 17% per 5-year interval, whereas the unspecified defect prevalence was stable until 1994 to 1995 and then declined (a 15% decrease per 5-year interval). A sex-specific difference in prevalence was only seen for severe defects; sex-specific prevalences were similar for septal defects, other specified defects, and unspecified defects. For severe defects, girls had a lower prevalence (23.3 per 10,000 [95% CI, 21.8-25.0]) than boys (31.5 per 10,000 [95% CI, 29.8-33.4]) before 1990, whereas girls and boys had similar prevalences from 1990 onward.
Figure 2.
Prevalence of CHD by birth cohorts for isolated septal defects, severe defects (heterotaxia, conotruncal defects, AVSD, APVR, LVOTO, RVOTO, complex defects, associations), other specified CHDs, and unspecified CHDs in 1,711,641 singletons born in Denmark—1977 to 2005.
Discussion
We identified and categorized congenital CHDs in >18,000 persons born in Denmark during a 29-year period, for an overall prevalence of 103 per 10,000 live births. The prevalence of CHDs increased from 73 to 113 per 10,000 live births during the study period—1977 to 2005. Among singletons, the prevalence of severe CHDs increased overall by 25%, whereas VSD prevalence doubled and ASD prevalence tripled. Increasing numbers of CHDs diagnosed in infancy contributed most to the increases in prevalence until 1996–1997, and thereafter, the prevalence became stable.
Congenital heart defects include a range of defects, that, according to errors in specific sequences of cardiac development,21 can be broadly be classified as looping defects, outlet defects, inlet defects, chamber development/ventricular septation defects, and endocushion tissue defects. Similar defects may result from errors in different sequences of cardiac development, or one defect may lead to a series of other defects (eg, Fallot tetralogy). Various heart malformations may occur in combination or as single defects.
Recently, Botto et al20 from the National Birth Defects Prevention Group in Atlanta, GA, published a system for classifying cardiac defects. They reviewed all available information on infants with CHDs registered in their hospital surveillance system and categorized the infants into >70 detailed cardiac phenotypes. Their procedure also contained a “mapping system” that allowed the detailed phenotypes to be grouped together into broader categories. In a similar approach using individual-level ICD codes, we constructed an algorithm to group certain combinations of ICD codes using a hierarchical procedure, yielding 17 cardiac phenotypes. A similar approach was used with ICD codes from administrative records in a study from Canada.9
Two events might have influenced case ascertainment and CHD classification. Outpatient hospital clinic diagnoses were first included in the National Patient Register in 1995, and we therefore expected some increase in the ascertainment of CHDs after 1995. However, the observed overall increase in birth prevalence of CHDs began 15 years before the registration of outpatient diagnoses was initiated, and CHD prevalence actually stabilized in the mid-1990s. Therefore, we assume that case ascertainment was virtually complete even before the registration of outpatient diagnoses began. Second, in 1994, the 10th version of the ICD was implemented in Denmark. ICD-10 includes more detailed congenital CHD coding than ICD-8, with the exception of codes for ASD and VSD. Consequently, although the shift from ICD-8 codes to ICD-10 codes cannot explain the increases in ASD and VSD prevalence observed in the late 1990s, it could explain the increasing prevalence of “other specified CHDs” and the decreasing prevalence of unspecified CHDs from 1994 onward.
The National Birth Defects Prevention Group in Atlanta has published CHD prevalences based on their classification scheme.22 The overall CHD prevalence in metropolitan Atlanta in the period 1998 to 2005 was 81.4 per 10,000 live births.22 In comparison, when PDAs in infants born preterm were excluded, our overall CHD prevalence was 93.2 per 10,000 live births in the period 1977 to 2005. Compared to the Atlanta group, we used a simpler approach to classification of CHDs, by using ICD codes and by reporting defects in broader categories.20, 22 Even so, most of our prevalence figures were comparable to those reported by the Atlanta group (eg, prevalences of ASD, AVSD, LVOTO, and heterotaxia). Compared to our study, the Atlanta group reported a lower prevalence of conotruncal defects, 7.9 per 10,000 live births, and higher prevalences of RVOTO, single ventricle (complex defects), VSD (7.0, 1.0, and 41.8 per 10,000 live births, respectively). Our RVOTO and complex defect prevalences may have been lower due to classification of some of these cases in the “unspecified defects” group, while our VSD prevalence was likely lower because our VSD group consisted of solely of isolated VSD (VSD without other concomitant heart defects). In addition, the Atlanta prevalences included defects ascertained among multiple births and stillbirths, whereas ours did not.
The overall increase in CHD prevalence in Denmark during the study period was primarily due to an increase in septal defects diagnosed during infancy. An “epidemic” of VSD was first described in an American study from 1980,14 with dozens of subsequent papers confirming the increase in VSD prevalence.5 The reason for an increase in septal defect prevalence over time in the present study is unknown, but improved detection of minor ventricular defects that later close spontaneously is a likely explanation. In a report from 2002, Hoffmann et al5 reviewed 62 studies reporting incidence of CHDs since 1955. Combining data from 44 studies, they found that the variation in incidence depended on the number of small ventricular defects reported. The incidence of moderate and severe CHDs was 6 per 1,000 live births, and with the addition of trivial defects, such as small ventricular septal defects, the incidence was 62 per 1,000 live births. The authors concluded that there was no evidence for a real change in CHD incidence over time. The increase in the prevalence of small defects coincided with improvements in ultrasound diagnostics beginning in the mid 1980s,23, 24 whereas the ascertainment of severe defects was less influenced by the availability of echocardiographic investigations.
We also found an increase in atrial septal defect prevalence, as shown in the Canadian study.9 The reason for the ASD increase is unclear, but the increase was probably also due to improved detection of such defects in infants, as described above for ventricular septal defects.
Among the severe defects, the prevalences of conotruncal defects, AVSD, and RVOTO increased significantly, whereas the prevalences of heterotaxia, APVR, and LVOTO were stable. It is unlikely that ascertainment of conotruncal defects, AVSD, and RVOTO improved over time, since severe CHDs almost always come to medical attention, either due to the need for surgery or at death. Changes in the distribution of CHD risk factors could have occurred and resulted in an increase in the prevalence of conotruncal defects, AVSD, or RVOTO. For example, the increasing prevalence of type 2 diabetes among women of childbearing age25 may also have contributed to the increase in CHD prevalence; diabetes has been shown to increase the risk of various types of CHD, with relative risks ranging from 3 to 57.8
An international collaborative study that included births with selected birth defects in the period 1968 through 199826 reported a male excess of left obstructive defects and transposition of the great arteries, a finding that was also reported from Canada9 and the Czech Republic.27 However, the overall prevalence of CHDs was higher among female children and adults in the Canadian study.9 In contrast, sex-specific prevalences did not differ significantly in our study, overall, for specific ages at diagnosis, or for septal or severe CHDs, although the percent increase in CHD prevalence tended to be higher for women than for men, particularly in the severe defects group.
The prevalence of CHDs diagnosed in infancy seemed to stabilize after 1996–1997, possibly because no further improvements in postnatal CHD diagnostics have occurred. However, an increase in prevalence due to improved diagnostics may also have been counteracted by increased prenatal screening for congenital malformations and subsequent termination of affected pregnancies. In fact, a study from France reported that overall birth defect prevalence has begun to decline because of increased induced abortion of affected fetuses.28 However, although prenatal detection of neural tube defect was high, the sensitivity for detecting CHDs was very low, meaning that a large proportion of CHDs cannot be avoided by detection and termination of affected pregnancies. Among all births with CHDs in the Danish county Odense, 1980 to 2006, the proportion of terminated pregnancies for fetal cardiac anomaly after prenatal investigation (prenatal chromosomal investigation and/or fetal echocardiography) was 3% (excluding PDA).19 Elective termination after prenatal investigation had a minor effect on the overall CHD birth prevalence in Denmark.
Our study had multiple strengths. Because our cohort encompassed the entire Danish population, with >1.7 million persons born during a 29-year period, our study had great overall power. Furthermore, Denmark's national registers allowed for complete follow-up of birth cohort members.29 The unique conditions in Denmark with free health care for all citizen and mandatory reporting of all diagnoses, registration of birth defects in Denmark is considered virtually complete. In the present study, prevalences of severe CHDs overall and by specific defect corresponded well with estimates from comparable population-based registers.12, 19
In conclusion, the increasing prevalence of CHDs reported internationally was confirmed in the present study and was found to be mostly due to an increasing number of isolated septal defects diagnosed in infancy, although we also found that severe defects contributed to the increase in CHD prevalence. The availability of echocardiographic investigations from the mid 1980s is likely to explain the increased detection of septal defects. Changes in pregnancy risk factors associated with CHDs could explain the rise in severe CHD prevalence. Finally, the stable prevalence of CHDs diagnosed in infancy after the period 1996–1997, could possible be due to no further improvements in postnatal CHD diagnostics.
Disclosures
The authors are solely responsible for the design and conduct of this study, all study analyses, and the drafting and editing of the manuscript and its final contents.
Acknowledgements
We thank Dr Lorenzo Botto for sharing his classification scheme with us, and Jan Hansen for providing the data from the Danish Cytogenetic Central Register.
Appendix A. Definition of cardiac phenotypes
Persons with heart defects were classified into 17 heart phenotypes by grouping specific ICD codes in hierarchical fashion. For example, all persons with any heterotaxia diagnosis, regardless of other heart defect codes, were allocated to the heterotaxia group. Next, all persons with conotruncal defect codes but without AVSD were identified; by definition, because of the hierarchical nature of the classification scheme, individuals with heterotaxia could not be included in the conotruncal defect group. Next, individuals with AVSD but without heterotaxia or conotruncal defects, were identified, and so on. Specifically, the phenotypes were defined as follows: heterotaxia with or without any other heart defect (ICD-10: Q24.0, Q24.1, Q89.3, Q20.6; ICD-8: 759.00, 759.01, 759.09); conotruncal defects [truncus arteriosus, interrupted aortic arch, d-transposition of great arteries, tetralogy of Fallot (TOF), double outlet right ventricle] without concomitant atrioventricular septal defect; conoventricular VSD without ASD or AVSD, pulmonary valve stenosis (PVS) and VSD, and pulmonary atresia and VSD without AVSD [Q20.0, Q25.1A, Q25.2, Q25.3, Q25.4, Q20.3, Q21.3, Q20.1, without Q21.2; (Q21.4, without Q21.1 or Q21.2); (Q22.1 and Q21.0, without Q21.2); (Q25.5 and Q21.0, without Q21.2); 746.09, 747.29, 746.19, 746.29, (746.63 and 746.39), (747.39 and 746.39), without 746.59]; AVSD (without TOF) [(Q21.2, without Q21.3), (746.59, without 746.29)]; APVR (total and partial APVR, but without AVSD) [Q26, Q26.2, Q26.4, Q26.8, Q26.9, without Q21.2; Q26.3, without Q21.2]; LVOTO [hypoplastic left heart syndrome, coarctation of aorta with intact ventricular septum, aortic stenosis] [Q23.4, (Q25.1, without Q21.0), Q23.0, Q23.1A; 746.79, (747.19, without 746.39), 746.62]; RVOTO [PVS only, tricuspid atresia, Ebstein anomaly, pulmonary artery atresia with intact ventricular septum and without TOF] [Q22.1 only, Q22.4, Q22.5, (Q25.5, without Q21.0 or Q21.3); 746.63 only, 746.61, (747.39, without 746.29)]; septal defects (VSD only, ASD only, VSD and ASD only) [Q21.0 only, Q21.1 only, (Q21.0 and Q21.1) only; 746.39 only, (746.40, 746.41, 746.49) only, 746.39 and (746.40, 746.41, 746.49) only]; complex defects (single ventricle) [Q20.4]; conotruncal defects with AVSD {[Q20.0, Q25.1A, Q25.2, Q25.3, Q25.4, Q20.3, Q21.3, Q20.1, (Q21.4 without Q21.1), (Q22.1 and Q21.0), (Q25.5 and Q21.0)] and Q21.2; [746.09, 747.29, 746.19, 746.29, (746.63 and 746.39), (747.39 and746.39)] and 746.59}; septal defects with LVOTO (VSD and coarctation of aorta, VSD/ASD and aortic stenosis, VSD/ASD and coarctation of aorta) [(Q21.0 and Q25.1), (Q21.1 and Q21.0 and Q23.0, Q23.1A), (Q21.1 and Q21.0 and Q25.1); (746.39 and 747.19), (746.39, 746.40, 746.41, 746.49 and 746.62), (746.39, 746.40, 746.41, 746.49 and 747.19)]; septal defects with RVOTO (ASD and PVS, VSD and PVS, VSD/ASD and PVS) [Q21.1 and Q22.1, Q21.0 and Q22.1, (Q21.1 and Q21.0 and Q22.1); (746.40, 746.41, 746.49) and 746.63, 746.39 and 746.63, (746.39, 746.40, 746.41, 746.49) and 746.63]; PDA (Q25.0 only, 747.09 only); unspecified [Q24.9 only; 746.89 only, 746.99 only]; and other specified heart defects.
References
- . Congenital heart disease in 56,109 births. Incidence and natural history. Circulation. 1971;43:323–332
- Estimating the prevalence of malformation of the heart in the first year of life using capture-recapture methods. Am J Epidemiol. 1999;150:778–785
- . Temporal variability in birth prevalence of cardiovascular malformations. Heart. 2000;83:414–419
- . Racial and temporal variations in the prevalence of heart defects. Pediatrics. 2001;107:e32
- . The incidence of congenital heart disease. J Am Coll Cardiol. 2002;39:1890–1900
- Temporal variability in birth prevalence of congenital heart defects as recorded by a general birth defects registry. J Pediatr. 2003;142:690–698
- Congenital cardiac malformations in Iceland from 1990 through 1999. Cardiol Young. 2004;14:396–401
- Non-inherited risk factors and congenital cardiovascular defects: Current knowledge: a scientific statement from the American Heart Association Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation. 2007;115:2995–3014
- Congenital heart disease in the general population: changing prevalence and age distribution. Circulation. 2007;115:163–172
- Heart Disease and Stroke Statistics–2007 Update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2007;115:e69–e171
- Evaluation of the General Practice Research Database congenital heart defects prevalence: comparison to United Kingdom national systems. Birth Defects Res A Clin Mol Teratol. 2007;79:309–316
- International Clearinghouse for Birth Defects. Surveillance and research. Available at: http://www.icbdsr.org. Last accessed June 23, 2008.
- Trends in hospital admissions, in-hospital case fatality and population mortality from congenital heart disease in England, 1994 to 2004. Heart. 2008;94:342–348
- Is there an epidemic of ventricular septal defects in the U.S.A.?. Lancet. 1980;Feb 23:407–408
- . Prevalence of congenital heart disease. Am Heart J. 2004;147:425–439
- . Births/Demographics 2005. CDC, US Department for Health and Human Services; 2008;Available at: http://www.cdc.gov/nchs/Last accessed May 9, 2008
- Genetic basis for congenital heart defects: current knowledge: a scientific statement from the American Heart Association Congenital Cardiac Defects Committee, Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation. 2007;115:3015–3038
- . Some epidemiological aspects of congenital heart disease in Denmark. Acta Paediatr Scand. 1980;69:619–624
- UK EUROCAT Central Registry. Prevalence data tables 1980-2005. Available at: http://www.eurocat.ulster.ac.uk/. Last accessed July 3, 2008.
- Seeking causes: Classifying and evaluating congenital heart defects in etiologic studies. Birth Defects Res A Clin Mol Teratol. 2007;79:714–727
- . Cardiac development. Oxford: Oxford University Press; 2007;
- Prevalence of congenital heart defects in Metropolitan Atlanta, 1998-2005. J Pediatr. 2008;
- . Increased prevalence of ventricular septal defect: epidemic or improved diagnosis. Pediatrics. 1989;83:200–203
- Increasing incidence of ventricular septal defects by improved detection rate. Acta Paediatr. 1994;83:653–657
- The continuing epidemics of obesity and diabetes in the United States. JAMA. 2001;286:1195–1200
- Sex and congenital malformations: an international perspective. Am J Med Genet. 2005;134A:49–57
- . Boy:girl ratio in children born with different forms of cardiac malformation: a population-based study. Pediatr Cardiol. 1994;15:53–57
- Impact of prenatal diagnosis on live birth prevalence of children with congenital anomalies. Ann Genet. 2002;45:115–121
- . Epidemiology. When an entire country is a cohort. Science. 2000;287:2398–2399
PII: S0002-8703(08)00907-1
doi:10.1016/j.ahj.2008.10.017
© 2009 Mosby, Inc. All rights reserved.
