Centers for Disease Control and Prevention
Centers for Disease Control and Prevention
Centers for Disease Control and Prevention CDC Home Search CDC CDC Health Topics A-Z    
Office of Genomics and Disease Prevention  
Office of Genomics and Disease Prevention

 

 Journal Publication

This paper was published with modifications in the American Journal of Medical Genetics 73:113-118 (1997)

Please note: Figures 1- 3 have not been included. Figure 1 contains a posterior view of an infant with craniorachischisis extending into the lumbar area. Figure 2 displays lateral and anterior views of an infant with iniencephaly. Figure 3 includes lateral and anterior views of an infant with craniorachischisis with spinal retroflextion.


Elevated Rates of Severe Neural Tube Defects in a High-Prevalence Area in Northern China

by Cynthia A. Moore, Song Li, Zhu Li, Shi-xin Hong, Hai-qin Gu, R.J. Berry, Joseph Mulinare, and J. David Erickson


bullet Abstract
bullet Introduction
bullet Materials and Methods
bullet Results
bullet Discussion
bullet Acknowledgments
bullet References
bullet Table I

Abstract

In the northern provinces of China, the birth prevalence rate of neural tube defects (NTDs) is among the highest in the world--at about 6 per 1,000 births in rural areas. A unique population-based birth defects surveillance system in which photographs are taken of infants with selected external birth defects was implemented in two provinces in northern China and two provinces in southern China where NTD rates approximate those in the United States. In the period from March 1992 through December 1993, 660 infants with NTDs were identified by the surveillance project from a birth cohort of 251,567.

We compared data from the two surveillance areas in China with data from a low-prevalence area in the United States to determine if the pattern of NTD types differs. Based on birth prevalence rates of NTDs from the Metropolitan Atlanta Congenital Defects Program, the observed to expected ratios for two types of NTDs were markedly increased at 80.8 for craniorachischisis and 25.0 for iniencephaly. Rates of these two NTDs in the southern provinces are increased to a lesser degree with observed to expected ratios of 7.1 for craniorachischisis and 2.7 for iniencephaly.

The pattern of NTDs in northern China shows an increase in types that are rare in low-prevalence areas such as metropolitan Atlanta. Increased awareness of varying patterns of NTDs in different populations may have important implications for identifying etiologic and pathogenetic mechanisms of NTDs.

Introduction

In the northern provinces of China the incidence of neural tube defects (NTDs) was reported previously in a hospital-based study to be about 7 per 1,000 births in rural areas [Lian et al., 1987]. This rate is among the highest in the world and is approximately ten-fold greater than rates in western surveillance systems [Centers for Disease Control and Prevention, 1993]. In China, NTDs are a major cause of stillbirth and infant mortality, accounting for about one-third of stillbirths and one-fourth to one-third of neonatal deaths [U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control, 1989]. To address this public health problem, the Centers for Disease Control and Prevention (CDC) has a cooperative agreement with Beijing Medical University (BMU) in China to evaluate a community intervention program using periconceptional folic acid supplementation to prevent NTDs.

One aspect of the program is its unique population-based birth defects surveillance system that uses photographs to document birth defects and provides baseline rates for NTDs and other external structural birth defects. This report compares the birth prevalence rates of the different types of NTDs occurring in two surveillance areas in China to those from a population-based surveillance system in the United States, the Metropolitan Atlanta Congenital Defects Program (MACDP).

Materials and Methods

The Birth Defects Surveillance for BMU-CDC Collaborative Project-China (BDSCP) was instituted in 11 cities and counties in two northern provinces, Hebei and Shanxi, in March of 1992. Surveillance activities were also begun in five cities and counties in two southern provinces, Jiangsu and Zhejiang, at the same time. An additional 16 southern cities and counties were added in January of 1993. During the period from March 1992 through December 1993, the project collected data on 3,766 infants with external structural birth defects from a birth cohort of 251,567 infants--90,709 infants from 11 cities and counties in the northern provinces and 160,858 infants from 21 cities and counties in the southern provinces. Folic acid supplementation was available only through a pilot project that encompassed approximately 10% of the women registering for marriage in one city and one county in the north.

The National Center for Maternal and Infant Health, BMU, had previously established a surveillance system, known as the Perinatal Health Care Delivery System, which monitors all pregnancies and their outcomes. Information on resident births is provided to the BDSCP monthly from each participating city and county. Data on infants with birth defects is collected by designated health care providers at the city and county levels. Live-born infants with birth defects are included if they are at least 20 weeks' gestational age and are ascertained as having a birth defect by six weeks of age. For each of these infants, basic demographic information (e.g., birth date, sex, birth order), a detailed clinical description of the birth defect(s), birth defect code(s), and photographs are collected. Photographs and basic demographic information are requested on all stillborn infants even if no structural anomaly is noted. Verbal consent to photograph is obtained from the parent(s) of the live-born or stillborn infant. In addition to didactic sessions on birth defects and clinical photography, the field staff is provided with a written reference, the Birth Defects Diagnosis Atlas, which was developed for this project. The atlas contains a detailed description, birth-defect code, photograph, and other information such as optimal photographic views for 26 common birth defects. Birth defects that are not included in the atlas are photographed, described, and then coded as "unknown" by the field staff.

All data on infants with birth defects are sent to BMU monthly, where they are reviewed by three pediatricians who independently re-code the defect. The initial code assigned by the field staff is the World Health Organization's 1979 International Classification of Diseases, 9th Revision, Clinical Modification [U.S. Department of Health and Human Services, Public Health Service, Health Care Financing Administration, 1980]. The BDSCP birth-defect code assigned at BMU is a minor modification of the MACDP birth-defect code. The code used by MACDP is based on ICD-9-CM and the British Paediatric Association's Classification of Diseases [British Paediatric Association, 1979]. A final review of diagnoses is done yearly by one of the authors (CAM) who is a pediatrician and clinical geneticist.

Neural tube defects are classified in the following manner in the BDSCP: a neural tube defect is considered to be "open" if either membranes or neural tissue are exposed; a "closed" lesion is skin-covered. Spina bifida refers to either an open or closed NTD confined to the spine. Anencephaly includes open defects affecting the cranium only and is divided into meroanencephaly (affecting the anterior cranium) and holoanencephaly (affecting the entire cranium, including the foramen magnum). Anencephaly with any contiguous involvement of the spine is classified as craniorachischisis (Fig 1). Iniencephaly is characterized by a closed cranium that is markedly retroflexed (Fig.2), and is classified into the following four groups: isolated iniencephaly, iniencephaly with encephalocele, iniencephaly with spina bifida, and iniencephaly with encephalocele and spina bifida. Iniencephaly is classified separately from craniorachischisis with retroflexion of the spine (Fig. 3) as described by Lemire et al. [1972]. The level of the lesion is designated by the sixth digit of the code for infants with spina bifida and craniorachischisis. For spina bifida the designation is for highest level of lesion; for craniorachischisis, it represents the lowest level of lesion. Encephalocele includes any saccular herniation, usually skin-covered, from the cranium. Encephalomyelocele is a similar herniation that spans a portion of the cranium and the spine. Cranioschisis refers to an open defect of the occiput, which is non-membrane covered.

We based the expected number of case infants in China on data from MACDP, a population-based birth defects surveillance system with active case ascertainment that operates in five counties of metropolitan Atlanta. All MACDP case infants have a serious congenital anomaly diagnosed in the first year of life, are either live born or stillborn at 20 weeks' gestation or greater, and have mothers who were residents of metropolitan Atlanta at the time of delivery. Full details of this surveillance system have been described elsewhere [Edmonds et al., 1981]. We used MACDP data from 1990 through 1994 in an attempt to have sufficient numbers of births that would include the rare types of NTDs. During this time there were 194,162 live births in Atlanta. We also adjusted the rates, increasing them by 30% in an attempt to compensate for the effects of prenatal diagnosis and subsequent pregnancy termination on NTD rates [Roberts et al., 1995; Centers for Disease Control and Prevention, 1995]. We calculated observed to expected ratios (O/E) and 95% confidence intervals (CI) using a statistical analysis package [James, 1990].

Results

The total NTD rate for the two northern provinces was 5.68 per 1,000 total births; in the two southern provinces the rate was 0.90 per 1,000 total births. Table I shows the number of case infants, birth prevalence rate per 1,000 births, and O/E with 95% CI for each of the most common forms of NTDs among the northern and southern provinces in China.

In the northern provinces the rates of all types of NTDs are significantly increased when compared with rates found in the MACDP data; however, the rates of craniorachischisis and iniencephaly are markedly increased with an O/E of 80.8 and 25.0, respectively. In the southern provinces of China, the rate of spina bifida is lower than the MACDP rate, whereas the rates of anencephaly and encephalocele are comparable in the two systems. Again, the rates of craniorachischisis and iniencephaly are increased in the Chinese population, but to a lesser extent with an O/E of 7.1 for craniorachischisis and 2.7 for iniencephaly. The degree of rachischisis among infants from both surveillance areas in China was severe with more than 40% extending into the infant's lumbar region. In addition, approximately 30% of infants with craniorachischisis had a marked degree of spinal retroflexion that resulted in a body habitus resembling iniencephaly. Two infants with the rare defect, cranioschisis, were also seen in the northern provinces; however, the rate of this NTD as well as encephalomyelocele cannot be compared to Atlanta data since there is no unique code for these defects in MACDP.

Discussion

The distribution of infants with the various types of NTDs in China is different than we expected; there are increased rates of NTDs that are usually considered rare ( i.e., craniorachischisis and iniencephaly), when compared with adjusted MACDP rates. This increase is more marked in the northern provinces but is still evident in the southern provinces where NTD rates approximate those seen in the United States.

Our adjustment of MACDP rates by 30% was conservative since percentages of NTD-affected pregnancies that were prenatally diagnosed and subsequently terminated varied from 20% to 70% in a recent multi-state study, depending on the type of NTD, the geographic location, and the race of the mother [Centers for Disease Control and Prevention, 1995]. This study and others [Limb and Holmes, 1994; Roberts et al., 1995] documented that pregnancies associated with anencephaly, a uniformly fatal NTD, were terminated more frequently than were those associated with spina bifida. Although this finding presumably would also hold for craniorachischisis and iniencephaly, there are no published rates for termination of pregnancies associated with these rare NTDs. A recent study of the NTD birth prevalence rate in MACDP for 1990 through 1991 showed that the decrease in the total NTD rate due to elective pregnancy termination following prenatal diagnosis was slightly greater than 30% [Roberts et al., 1995], we used this overall rate for adjustment in the current study.

Several authors have noted that rates of iniencephaly and anencephaly (or anencephaly with spina bifida) were higher in areas with high overall NTD rates [Warkany,1971; David and Nixon, 1976]; however, it is difficult to compare rates as determined in different studies because of variations in the way that NTDs are reported. For example, anencephaly and craniorachischisis are often tallied together, or craniorachischisis is coded as two defects. Iniencephaly is also frequently included with anencephaly. One study from EUROCAT attempted to separate the different types of NTDs that occurred in Europe for the period 1980-1987 [Dolk et al., 1991]. Although the higher-incidence areas in the United Kingdom and Ireland had an overall lower rate of NTDs at about 3 per 1,000 births than northern China, these rates were still two to three times higher than those in continental Europe and Malta. The same pattern of increased birth prevalence rates for craniorachischisis and iniencephaly were seen in the United Kingdom and Ireland. Unfortunately, the researchers were unable to differentiate contiguous from non-contiguous lesions in all infants with anencephaly and spina bifida. They were able to classify iniencephaly and craniorachischisis with spinal retroflexion separately; however, in their study, these NTDs are called iniencephaly and anencephaly with iniencephaly, respectively.

This report is the first documentation of high rates of craniorachischisis and iniencephaly, two NTDs that are almost invariably fatal, in an area of northern China with a high birth prevalence of NTDs. The data are population-based and include large numbers of births. There is a photographic record of approximately 80% of the infants with birth defects, and in the field, case ascertainment requires limited diagnostic expertise.

The data also have limitations. About 4% of the initial photographs were technically poor or had inadequate views to make a diagnosis. Approximately 20% of infants were not photographed, and only a written description of the defect was available. We think that most infants with NTDs who were not photographed could be correctly classified as to type of NTD since the descriptions were extensive; however, it is possible that some were misclassified. For example, iniencephaly without encephalocele or spina bifida may have been reported as hydrocephaly if the neck retroflexion was not appreciated. With this system, only external anomalies are captured, and there is no radiographic or pathologic validation of the defect. It is therefore not possible to obtain some types of specific information about the NTD, such as the exact level or type of spina bifida. The upward 30% correction of MACDP NTD rates because of prenatal diagnosis and subsequent pregnancy termination may have underestimated the effect of such elective terminations on rates of the severe types of NTDs and overestimated this effect on rates of the less severe types.

The finding of elevated rates of both craniorachischisis and iniencephaly in high-incidence areas such as northern China is interesting. Whether or not these two types of NTDs should be classified together, especially craniorachischisis with spinal retroflexion and iniencephaly, has been debated [Howkins and Lawrie, 1939; Coffey, 1965; Lemire et al., 1972]. Because the presence or absence of a neurocranium suggests different times of onset for these two malformations, Lemire and colleagues [1972] suggest separate classification. Blackburn and coworkers [1995] found major differences in the spinal malformations in these two NTDs; they concluded that the cervical spine duplication in iniencephaly pointed to possible notochord-mesenchymal dysfunction rather than nonclosure of the neural tube as an etiology for iniencephaly.

The inheritance of NTDs is usually described as being multifactorial or the result of the interaction of many genes and many environmental factors. Assumptions underlying the multifactorial threshold model, which is used to predict recurrence risks in families, include higher underlying genetic liability among more severely affected individuals and a lower genetic threshold in populations with higher rates [Carter, 1970]. It would be interesting to test these assumptions in the northern Chinese population, but unfortunately, recurrence data for NTDs are not yet available.

In conclusion, data from this unique population-based surveillance program in China show an unusual distribution of NTD types in northern China compared with the pattern typically found in lower rate areas. A similar pattern was documented in at least one other high-incidence area in the United Kingdom and Ireland; however, lack of uniform classification or incomplete classification in most epidemiologic studies of NTDs makes comparisons of patterns difficult. The finding of an elevated O/E for craniorachischisis and iniencephaly in southern China where the NTD rate is lower and similar to rates seen in western surveillance systems is interesting, and suggests that factors which lead to higher birth prevalence rates may not be solely responsible for the pattern.

Identifying these unusual patterns of NTDs in different populations may be an important step in identifying etiologic and pathogenetic mechanisms. These data raise several questions about NTDs in high-incidence populations, including 1) What is the cause of the unusual pattern of NTDs? 2) Is there a common developmental mechanism for craniorachischisis and iniencephaly? and 3) What will be the response to intervention efforts such as folic acid supplementation?

Acknowledgments

We thank the China-US Collaborative Project for NTD Prevention, which includes on the Chinese side: Department of Science and Technology, Chinese Ministry of Health; National Center for Maternal and Infant Health, Beijing Medical University; Department of Health Care Epidemiology, School of Public Health, Beijing Medical University; and government, public health, and maternal and child health workers from Hebei, Shanxi, Zhejiang, and Jiangsu provincial, county, township, and village levels. On the United States side: Birth Defects and Genetic Diseases Branch, Division of Birth Defects and Developmental Disabilities, and Nutritional Biochemistry and Molecular Biology Branches, Division of Environmental Health Laboratory Sciences, National Centers for Environmental Health. We appreciate the assistance of the WHO Collaborating Center in Perinatal Care and Health Services Research in Maternal and Child Health, Atlanta, Georgia. We thank Mr. Yecai Liu, Ms. Mei Wang, Ms. Lan Xiao and Ms. Zhihui Gu for their efforts in the maintenance of the BDSCP database; Dr. Ron Lemire for assistance in classification of infants with craniorachischisis and iniencephaly; Mr. Lee James for statistical assistance; and Dr. Muin Khoury for analytic suggestions. Finally, we thank the MACDP abstractors, Jo-Anne Croghan, Joan Garcia, Debbie Nurmi, Joann Garcia, Charlie Mae Peters, Melissa Johnson, and Carolyn Sullivan.

References

  1. Blackburn W, Maertens P, Cooley NR Jr., Barr M Jr. (1995): Studies of the cranioskeletal anomalies associated with iniencephaly. Proceedings of the Greenwood Genetic Center 14:87.
  2. British Paediatric Association (1979): British Paediatric Association Classification of Diseases: Codes Designed for Use in the Classification of Paediatric and Perinatal Disorders. London: British Paediatric Association, Vol. 2.
  3. Carter CO (1970): Multifactorial genetic disease. Hosp Pract 5:45-59.
  4. Centers for Disease Control and Prevention (1993): Congenital malformations surveillance. Teratology 48:545-700.
  5. Centers for Disease Control and Prevention (1995): Surveillance for anencephaly and spina bifida and the impact of prenatal diagnosis--United States, 1985-1994. MMWR 44:1-13.
  6. Coffey VP (1965): Anencephaly or iniencephaly? J Irish Med Assoc 56:96.
  7. David TJ, Nixon A (1976): Congenital malformation associated with anencephaly and iniencephaly. J Med Genet 13:263-265.
  8. Dolk H, De Wals P, Gillerot Y, Lechat MF, Ayme S, Cornel M, Cuschieri A, Garne E, Goujard J, Laurence KM, Lillis D, Lys F, Nevin N, Owens J, Radic A, Stoll C, Stone D, Ten Kate L: Heterogeneity of neural tube defects in Europe: the significance of site of defect and presence of other major anomalies in relation to geographic differences in prevalence. Teratology 44:547-559.
  9. Edmonds LD, Layde PM, James LM, Flynt JW, Erickson JD, Oakley GP (1981): Congenital malformation surveillance: two American systems. Int J Epidemiol 10:247-252.
  10. James LM (1990): "Statistical Analysis for Epidemiologic Research (SABER)." Atlanta, Georgia: Centers for Disease Control.
  11. Howkins J, Lawrie RS (1939): Iniencephalus. J Obstet Gynaecol Br Comm 46:35-31.
  12. Lemire RJ, Beckwith JB, Shepard TH (1972): Iniencephaly and anencephaly with spinal retroflexion. A comparative study of eight human specimens. Teratology 6:27-36.
  13. Lian Z-H, Yang H-Y, Li Z (1987): Neural tube defects in Beijing-Tianjin area of China. Urban-rural distribution and some other epidemiological characteristics. Journal of Epidemiology and Commununity Health 41:259-262.
  14. Limb CJ, Holmes LB (1994): Anencephaly: changes in the prenatal detection and birth status, 1972 through 1990. Am J Obstet Gynecol 13:9-16.
  15. Roberts HE, Moore CA, Cragan JE, Fernhoff PM, Khoury MJ (1995): Impact of prenatal diagnosis on the birth prevalence of neural tube defects, Atlanta, 1990-1991. Pediatrics 96:880-883.
  16. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control (1989): The Risk Approach in Perinatal Health. Shunyi County, People’s Republic of China. Washington, DC: Government Printing Office, DHHS publication no. 890-8412.
  17. U.S. Department of Health and Human Services, Public Health Service, Health Care Financing Administration (1980): ICD-9-CM: The International Classification of Diseases, 9th Revision, Clinical Modification, 2nd Ed. Washington, DC: Government Printing Office, DHHS publication no. (PHS) 80-1260.
  18. Warkany J (1971): Iniencephaly. In "Congenital Malformations." Chicago: Year Book Medical Publishers, Inc., pp. 234-236.

Table I. Observed to expected ratios (O/E) and 95% confidence intervals (CI) for selected neural tube defect types between surveillance areas in northern and southern China, 1992-1993, and Atlanta.

  Atlanta* China
  North South North South
Type of NTD Adj. RateH Cases RateH Cases RateH O/E 95% CI O/E 95% CI
Anencephaly 0.26 84 0.93 60 0.37 3.6 (2.8,4.4) 1.4 (1.1,1.8)
Spina Bifida 0.49 265 2.92 42 0.26 5.9 (5.2,6.7) 0.5 (0.4,0.7)
Craniorachischisis 0.01 97 1.07 15 0.09 80.8 (65.5,98.6) 7.1 (4.0,11.8)
Iniencephaly 0.01 15 0.16 3 0.02 25.0 (14.0,41.2) 2.7 (0.6,8.0)
Encephalocele 0.14 51 0.56 24 0.15 4.0 (5.7,6.8) 1.1 (0.7,1.6)
TotalI 0.91 516 5.68 144 0.90 6.2 (5.7,6.8) 1.0 (0.8,1.2)

*MACDP data from 1990-1994, adjusted upward by 30%; H Rate per 1,000 births; I Totals include 2 Chinese infants with cranioschisis and 19 Chinese infants with an encephalomyelocele. These defects cannot be discriminated as distinct NTD types in MACDP. Thirteen Chinese infants have two types of NTDs and are counted as a case infant in each category, but each case infant is counted only once in the total.MACDP data from 1990-1994, adjusted upward by 30%; H Rate per 1,000 births; I Totals include 2 Chinese infants with cranioschisis and 19 Chinese infants with an encephalomyelocele. These defects cannot be discriminated as distinct NTD types in MACDP. Thirteen Chinese infants have two types of NTDs and are counted as a case infant in each category, but each case infant is counted only once in the total.