SEARCH

SEARCH BY CITATION

Keywords:

  • anencephaly;
  • epideimology;
  • neural tube;
  • myelomeningocele;
  • spina bifida

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PREVALENCE
  5. HETEROGENEITY
  6. RISK FACTORS
  7. CONCLUSIONS
  8. REFERENCES

The epidemiological investigation of the common open neural tube defects (NTDs), anencephaly, and spina bifida, has a long history. The most significant finding from these past studies of NTDs was the identification of the protective effect of maternal, periconceptional supplementation with folic acid. Fortuitously, the association between folic acid and NTDs became widely accepted in the early 1990s, at a time when genetic association studies of complex traits were becoming increasingly feasible. The confluence of these events has had a major impact on the direction of epidemiological, NTD research. Association studies to evaluate genes that may influence the risk of NTDs through their role in folate-related processes, or through other metabolic or developmental pathways are now commonplace. Moreover, the study of genetic as well as non-genetic, factors that may influence NTD risk through effects on the nutrient status of the mother or embryo has emerged as a major research focus. Research efforts over the past decade indicate that gene–gene, gene–environment, and higher-order interactions, as well as maternal genetic effects influence NTD risk, highlighting the complexity of the factors that underlie these conditions. The challenge for the future is to design studies that address these complexities, and are adequately powered to detect the factors or combination of factors that influence the development of NTDs. © 2005 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PREVALENCE
  5. HETEROGENEITY
  6. RISK FACTORS
  7. CONCLUSIONS
  8. REFERENCES

Neural tube defects (NTDs) are generally believed to result from failure of fusion of the neural tube during early embryogenesis. The most common NTDs are anencephaly, which results from failure of fusion of the cranial neural tube, and myelomeningocele (commonly called spina bifida), which results from failure of fusion in the spinal region of the neural tube. Anencephaly and myelomeningocele are often referred to as “open” NTDs because the affected region of the neural tube is exposed to the body surface. Craniorachischisis, which results from failure of fusion of the neural tube over the entire body axis, is an additional, relatively rare form of open NTD.

In addition to the open NTDs, there are also a number of “closed” or skin-covered conditions that involve the neural tube, including: encephalocele, meningocele (also called closed spina bifida), iniencephaly, lipomeningocele, sacral agensis, and occult spinal dysraphisms (also referred to as spina bifida occulta) [Lemire, 1988]. The relationship between these skin-covered conditions and open NTDs has not been firmly established. However, families segregating both open and closed malformations have been reported (for example, [Fellous et al., 1982; Fineman et al., 1982; Moore et al., 1990]), and a decline in the prevalence of encephalocele, as well as anencephaly and spina bifida, was reported following folic acid fortification of the food supply in Chile [Castilla et al., 2003]. These observations suggest that there may be an overlap in the factors that influence the development of open and closed NTDs.

Epidemiological investigations of NTDs have tended to focus on anencephaly and spina bifida (i.e., myelomeningocele). The focus on these two conditions is likely to be explained by their prevalence relative to other NTDs, as well as their dramatic clinical appearance and consequences. Anencephaly is a lethal condition and, in the absence of adequate medical and surgical interventions, mortality rates for spina bifida also approach 100%. With appropriate treatment, the majority of individuals with spina bifida will survive. Recent population-based data from Atlanta, Georgia (USA) indicate that survival among individuals with spina bifida is approximately 87% at 1 year, and 78% at 17 years [Wong and Paulozzi, 2001]. However, affected individuals are at risk for a range of physical and developmental disabilities [Althouse and Wald, 1980; Casari and Farrall, 1998; Buccimazza et al., 1999; Singhal and Mathew, 1999; Bowman et al., 2001], and experience excess mortality into the adult years [Hunt, 1999; Singhal and Mathew, 1999; McDonnell and McCann, 2000; Bowman et al., 2001].

PREVALENCE

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PREVALENCE
  5. HETEROGENEITY
  6. RISK FACTORS
  7. CONCLUSIONS
  8. REFERENCES

Geographic and temporal variation in the prevalence of anencephaly and spina bifida is well documented [Olney and Mulinare, 1998, 2002]. In addition, within a given geographic location and time period, the prevalence of these conditions varies by race and ethnicity. For example, in California, the prevalence of NTDs was reported to be highest in Hispanics (1.12/1,000), lowest in Blacks and Asians (0.75/1,000), and intermediate in non-Hispanic Caucasians (0.96/1,000) [Feuchtbaum et al., 1999].

The relative contribution of genetic and environmental (e.g., social, dietary) differences between racial and ethnic groups, to the observed differences in NTD prevalence, has not been established.

The relative contribution of genetic and environmental (e.g., social, dietary) differences between racial and ethnic groups, to the observed differences in NTD prevalence, has not been established.

Some studies of Mexican-Hispanics residing in the United States have reported that the risk of NTDs is higher in the offspring of women who were born in Mexico than in those born in the United States [Shaw et al., 1997; Hendricks et al., 1999]. Such observations suggest that environmental differences are likely to contribute to the variation in NTD prevalence between groups. However, other studies in Mexican-Hispanic populations have found no relationship between maternal birth place (Mexico vs. United States) and the risk of having a child with a NTD [Canfield et al., 1996].

The birth prevalence of NTDs is influenced by the availability of prenatal diagnosis and elective pregnancy termination [Chan et al., 1993; CDC, 1995; Velie and Shaw, 1996; Palomaki et al., 1999; Forrester and Merz, 2000; Rankin et al., 2000]. Termination rates are generally higher for anencephalic than for spina bifida affected fetuses [Chan et al., 1993; CDC, 1995; Velie and Shaw, 1996; Palomaki et al., 1999; Forrester and Merz, 2000; Rankin et al., 2000]. In addition, termination rates exhibit both temporal [Rankin et al., 2000] and regional [CDC, 1995] variation. These factors complicate surveillance efforts, since electively terminated fetuses with NTDs must be ascertained in order to monitor trends in NTD prevalence across or between populations, and to reduce the effects of selection bias in studies of potential risk factors [Borman and Cryer, 1990; Velie and Shaw, 1996].

A woman's risk of having a child with a NTD can be significantly reduced by periconceptional, folic acid supplementation [MRC, 1991; Czeizel and Dudas, 1992]. However, public health campaigns promoting daily use of multivitamins with folic acid have not had an appreciable impact on the prevalence of NTDs [Abramsky et al., 1999; Olney and Mulinare, 2002]. In contrast, data from countries that have implemented mandatory folic acid food fortification programs indicate a 30%–50% reduction in the prevalence of NTDs post-fortification [Honein et al., 2001; Erickson, 2002; Persad et al., 2002; Ray et al., 2002; Castilla et al., 2003; De Wals et al., 2003; Mersereau et al., 2004].

However, public health campaigns promoting daily use of multivitamins with folic acid have not had an appreciable impact on the prevalence of NTDs. In contrast, data from countries that have implemented mandatory folic acid food fortification programs indicate a 30%–50% reduction in the prevalence of NTDs post-fortification.

HETEROGENEITY

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PREVALENCE
  5. HETEROGENEITY
  6. RISK FACTORS
  7. CONCLUSIONS
  8. REFERENCES

Syndromes

NTDs can occur as part of malformation syndromes resulting from known chromosomal abnormalities (e.g., trisomy 13, 18, and 21) and single gene disorders (e.g., Meckel-Gruber and Waadenburg syndromes) [Hall and Solehdin, 1998]. In addition, rare families, segregating anencephaly and/or spina bifida (with or without other forms of NTDs) in patterns consistent with X-linked and autosomal recessive inheritance have also been reported [Baraitser and Burn, 1984; Toriello and Higgins, 1985; Farag et al., 1986; Jennson et al., 1988]. However, the vast majority of cases cannot be attributed to either chromosomal aberrations or the effects of a single genetic locus.

In a small proportion of cases, anencephaly and spina bifida occur as part of malformation syndromes that result from teratogenic exposures. The teratogenic potential of maternal pre-gestational diabetes is well established and includes a two- to tenfold increase in the risk of central nervous system malformations (including NTDs) among the offspring of affected women, relative to the general population [McLeod and Ray, 2002]. In addition, maternal use of valproic acid and/or carbamazepine is associated with an increased risk of spina bifida [Lammer et al., 1987; Hernandez-Diaz et al., 2001; Matalon et al., 2002]. Among women taking these medications, the risk of having a pregnancy affected with spina bifida may be as high as 1%–2% [Koren, 1999], whereas the risk of anencephaly does not appear to be increased.

In addition, maternal use of valproic acid and/or carbamazepine is associated with an increased risk of spina bifida. Among women taking these medications, the risk of having a pregnancy affected with spina bifida may be as high as 1%–2%, whereas the risk of anencephaly does not appear to be increased.

Malformation Level

In studies of the non-syndromic forms of anencephaly and spina bifida, these two conditions are often considered as a single entity. The legitimacy of such grouping is supported on embryological grounds (i.e. anencephaly and spina bifida are both neural tube closure defects [Lemire, 1988]), as well as by family studies that demonstrate that the sibs of individuals with spina bifida or anencephaly are at increased risk for both spina bifida and anencephaly (e.g., [Laurence et al., 1967; Carter and Evans, 1973; Nevin and Johnston, 1980]), and other similarities in the characteristics of these two conditions (e.g., the prevalence of both anencephaly and spina bifida is increased in Hispanics compared to non-Hispanic Caucasians [Canfield et al., 1996; Shaw et al., 1997]). However, there is also evidence that the factors influencing the development of anencephaly and spina bifida may not overlap completely. For example, although anencephaly and spina bifida co-occur within families, affected sib pairs are more likely to be concordant (e.g., anencephaly–anencephaly) than discordant (e.g., anencephaly–spina bifida) [Laurence et al., 1967; Carter and Evans, 1973; Nevin and Johnston, 1980]. In addition, exposure to valproic acid is associated with an increased risk of spina bifida, but not anencephaly.

It has also been suggested that there may be heterogeneity in the factors that influence the development of “upper” and “lower” NTDs. This subdivision has been justified on embryological grounds: the upper portion of the neural tube is formed by folding and fusion (i.e., neurulation), whereas the lower portion is formed by a different process (i.e., canalization). During the 1980s and early 1990s several studies that evaluated concordance for upper and lower NTDs among related individuals were published. These studies based their definitions of upper and lower defects on work that suggested canalization was responsible for neural tube development posterior to the 25th somite. Some of these studies reported complete concordance for upper or lower defects among affected relatives [Toriello and Higgins, 1985; Hall et al., 1988], whereas other studies reported both concordant and discordant affected relatives [Frecker et al., 1988; Seller, 1990; Drainer et al., 1991; Torok and Papp, 1991]. The latter studies indicate that the causes of upper and lower defects (as defined in these studies) are likely to be related. However, these studies may not have truly addressed the issue of heterogeneity in the causes of neurulation and canalization defects, since there is evidence that canalization is limited to a more distal segment of the neural tube (i.e., posterior to somites 32–34) [Nievelstein et al., 1993; O'Rahilly and Muller, 1999] than was assumed.

Associated Malformations

It has also been suggested that there may be etiological differences between NTDs that occur with and without additional malformations (after exclusion of cases with known chromosomal, genetic or teratogenic syndromes) [Khoury et al., 1982a,b]. However, the presence of additional malformations appears to be related to the location of the NTD [Seller and Kalousek, 1986; Davies and Duran, 2003], making it difficult to determine which of these characteristics may serve as a better indictor for the subdivision of cases into etiologically more homogenous subgroups. Given that anencephaly and spina bifida, as well as upper and lower NTDs co-segregate within families, it would appear reasonable to study all open NTDs as a single group. However, it would also seem prudent to perform subgroup analyses based on lesion level and/or the presence of additional malformations when sample sizes permit.

RISK FACTORS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PREVALENCE
  5. HETEROGENEITY
  6. RISK FACTORS
  7. CONCLUSIONS
  8. REFERENCES

The epidemiological characteristics of anencephaly and spina bifida have been extensively studied, and numerous review articles have been published [Mitchell, 1997; Olney and Mulinare, 1998; Botto et al., 1999; Frey and Hauser, 2003]. Although a number of potential risk factors for NTDs have been identified, many of the reported associations have been weak and have not been consistently replicated. Maternal obesity has emerged as a consistent risk factor for NTDs, with women in the highest body mass index categories (usually defined as a pre-pregnancy body mass index greater than 29 kg/m2) having a 1.5- to 3.5-fold higher risk than women with lower indices [Waller et al., 1994; Shaw et al., 1996, 2000; Watkins et al., 1996, 2003; Werler et al., 1996; Hendricks et al., 2001].

Maternal obesity has emerged as a consistent risk factor for NTDs, with women in the highest body mass index categories having a 1.5- to 3.5-fold higher risk than women with lower indices.

There is also relatively strong evidence that maternal hyperthermia increases the risk of having a child with an NTD by up to twofold [Milunsky et al., 1992; Lynberg et al., 1994; Lapunzina, 1996; Chambers et al., 1998; Shaw et al., 1998]. Other exposures that are currently of interest, but which require further investigation to establish their relationships with NTDs, include: food contaminated with fumonisins (a class of myocotoxins) [Hendricks, 1999; Sadler et al., 2002], chlorination disinfection by-products in drinking water [Klotz and Pyrch, 1999; Dodds and King, 2001], electromagnetic fields [Blaasaas et al., 2002], hazardous waste sites [Dolk et al., 1998; Orr et al., 2002], pesticides [Shaw et al., 1999], and maternal stress and social support [Carmichael et al., 2003a; Suarez et al., 2003a].

Undoubtedly, the most significant epidemiological finding with respect to NTDs is the protective effect of maternal periconceptional folic acid supplementation. This finding has been translated into public health policies, including educational campaigns and food fortification programs. Initial reports from countries that have implemented fortification programs indicate that the population prevalence of NTDs has declined following fortification [Honein et al., 2001; Erickson, 2002; Persad et al., 2002; Ray et al., 2002; Castilla et al., 2003; De Wals et al., 2003; Mersereau et al., 2004]. Hence, folic acid fortification of the food supply appears to represent the first successful, population-based, strategy for the primary prevention of a common congenital malformation.

Hence, folic acid fortification of the food supply appears to represent the first successful, population-based, strategy for the primary prevention of a common congenital malformation.

Although evidence for an association between NTDs and folic acid accumulated during the 1970s and 1980s, it was the results of the MRC [MRC, 1991] and Hungarian trials [Czeizel and Dudas, 1992] of folic acid suplementation that fueled public health efforts to reduce the prevalence of NTDs. Fortuitously, publication of the trial results occurred at a time when genetic studies of complex, non-Mendelian conditions, such as NTDs, were becoming increasingly feasible. In combination, these events have had a substantial influence on the direction of epidemiological studies on NTDs.

Given the observed association between folic acid and NTDs, it is not surprising that there has been substantial interest in the relation between genes involved in folate-related metabolic pathways and NTDs. Variants of several such genes have been found to be significantly associated with the risk of NTDs in one or more studies [Finnell et al., 2003]. Moreover, there is evidence that folate-related genes may exert their influence on the risk of NTDs via gene–gene [Wilson et al., 1999; De Marco et al., 2003; Zhu et al., 2003] and gene–environmental [Shaw et al., 2002; Volcik et al., 2003] interactions, and that such genes may act via either the maternal [Doolin et al., 2002] or the embryonic [Shields et al., 1999] genotype. However, firm conclusions regarding the relationship between NTDs and variants of specific folate-related genes are largely precluded at this time.

In addition to genes involved in folate-related metabolic pathways, there is currently considerable interest in genes that may influence the risk of NTDs through other metabolic or developmental pathways. Studies to date have not established any such gene as a risk factor for NTDs [Harris, 2001]. However, in general, the relationship between NTDs and “non-folate” genes has not been adequately studied.

There is also an interest in the relationship between NTDs and nutrients, and factors that effect nutrients, other than folic acid. There is an accumulating body of literature that supports an association between maternal vitamin B12 levels and NTDs [Ray and Blom, 2003; Suarez et al., 2003b].

There is also an interest in the relationship between NTDs and nutrients, and factors that effect nutrients, other than folic acid. There is an accumulating body of literature that supports an association between maternal vitamin B12 levels and NTDs.

In addition, recent studies have suggested potential associations between NTDs and maternal myo-inositol, zinc, and glucose levels [Groenen et al., 2003a,b], maternal intake of sucrose and foods with high glycemic index values [Shaw et al., 2003], maternal dieting behavior [Carmichael et al., 2003b] and physical activity [Carmichael et al., 2002], and maternal diarrhea [Felkner et al., 2003].

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PREVALENCE
  5. HETEROGENEITY
  6. RISK FACTORS
  7. CONCLUSIONS
  8. REFERENCES

Although the identification of the relationship between folic acid and NTDs should be heralded as one of the great successes of epidemiological research, the identification of additional NTD risk factors has proven to be exceedingly difficult. Interestingly, the initial experience with genetic association studies of NTDs has largely mirrored the experience with association studies of environmental risk factors for NTDs, that is the majority of the reported genetic associations have been weak and difficult to replicate. This relative lack of success has emerged as a characteristic common to genetic association studies of complex, human traits [Hirschhorn et al., 2002], and can be attributed to several factors including: (a) small, underpowered studies, (b) narrowly focused studies that consider only one or a few variants per gene, (c) lack of independent studies, (d) non-replication in independent studies, (e) failure to adequately address the influence of both the maternal and embryonic genotype, and (f) failure to adequately address potential effect modifiers.

The somewhat disappointing track record of genetic association studies of complex diseases has led some to question the utility of this approach [Gambaro et al., 2000; Holtzman, 2001; Strohman, 2002]. However, others have provided compelling arguments that genetic association studies can be informative when they are appropriately designed and analyzed [Ioannidis et al., 2003; Lohmueller et al., 2003; Redden and Allison, 2003; Zondervan and Cardon, 2004]. The challenge now is to design studies that minimize the potential for confounding and bias, and are adequately powered to detect the associations of interest. More than ever, this will require the collaboration of epidemiologists and geneticists, as well as individuals with expertise in nutrition, biochemistry, and other areas of the biomedical sciences.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PREVALENCE
  5. HETEROGENEITY
  6. RISK FACTORS
  7. CONCLUSIONS
  8. REFERENCES