Global Birth Prevalence of Spina Bifida by Folic Acid Fortification Status: A Systematic Review and Meta-Analysis

PLAIN-LANGUAGE SUMMARY

We sought to review the international prevalence of spina bifida. We conducted a systematic review in duplicate by using terms related to spina bifida and epidemiology. We performed meta-analysis stratified by fortification status, birth population, and geographic region. Overall, we identified 4078 articles, with 179 studies included in the systematic review and 123 studies included in a meta-analysis. In studies of live births (LBs), period prevalence estimates of spina bifida from 1985 to 2010 were lower in regions with mandatory (33.86 per 100 000 LBs) versus voluntary (48.35 per 100 000 LBs) folic acid fortification. Period prevalence estimates of spina bifida were also lower in studies of LBs, stillbirths, and terminations of pregnancy with mandatory (35.22 per 100 000 LBs) compared with voluntary (52.29 per 100 000 LBs) fortification. In LBs, stillbirths, and terminations of pregnancy studies, the lowest pooled prevalence estimate was in North America (38.70 per 100 000 LBs). Disparities in spina bifida prevalence remain between countries with and without mandatory folic acid fortification. Even in countries with mandatory folic acid fortification, studies restricted to LBs may underestimate the prevalence of spina bifida and prevalence estimates from LBs only should be interpreted with caution.

Birth defects are one of the leading causes of infant mortality worldwide^1–3 and affect an estimated 1% to 3% of all births.⁴ The etiology of many birth defects remains unknown despite a high prevalence and this reality hinders primary prevention efforts. A notable exception to this is the widespread decline in the prevalence of neural tube defects (anencephaly, spina bifida, encephalocele) following the mandatory fortification of grain products with folic acid in several countries.^5–7 Guidelines from the 1990s suggest a daily intake of 400 micrograms for women of reproductive age.⁸ Public health messaging aimed at increasing folate consumption has not had the same effect as mandatory folic acid fortification on increasing serum folate levels and reducing the birth prevalence of neural tube defects.⁹

The term spina bifida encompasses a group of birth defects (meningocele, meningomyelocele, myelocele, myelomeningocele, and rachischisis) that are the result of an incomplete closure of the spinal column leading to a herniation or exposure of the spinal cord or meninges.¹⁰ Although spina bifida has a lower case fatality rate than other neural tube defects (approximately 7%, compared with 46% for encephalocele, and 100% for anencephaly),¹¹ it can result in severe life-long morbidity.^12,13 This likely contributes to the high pregnancy termination rates following prenatal detection of spina bifida.^9,14

The predictive value of prenatal screening to detect spina bifida and other neural tube defects varies over time and across jurisdictions.¹⁵ An Australian study conducted in the late 1980s concluded that ultrasound was able to detect 75% of pregnancies with spina bifida, and serum screening for α-fetoprotein was able to detect 63% of affected pregnancies, resulting in a combined detection rate of 76% (95% confidence interval [CI] = 69%, 84%).¹⁵ More recent European data suggest that 88% (95% CI = 86%, 90)% of neural tube defects are prenatally diagnosed.¹⁴ Prenatal detection of open lesions (approximately 75% of spina bifida cases) typically has higher detection rates than closed lesions.¹⁵

Many countries have yet to mandate folic acid fortification of their grain products¹⁶ despite decades-old evidence that dietary supplementation of folic acid significantly decreases cases of neural tube defects, and spina bifida, in particular.¹⁷ Significant gains have been made in reducing child mortality rates, yet neonatal mortality reductions remain elusive because of a paucity of data and wide variations in neonatal health surveillance methods.^18–21 This systematic review provides an overview of international prevalence estimates for spina bifida, a condition still associated with significant childhood morbidity.

The objective of this study was to determine the global prevalence of spina bifida in live births (LBs), stillbirths, and terminations of pregnancy (TOPs) accounting for differential folic acid fortification policies, and to describe regional differences in prevalence.

METHODS

We performed this systematic review and meta-analyses with a well-known protocol (Preferred Reporting Items for Systematic Reviews and Meta-Analyses [PRISMA]).²² The study team generated the search strategy (Appendix A, available as a supplement to the online version of this article at http://www.ajph.org) with input from a research librarian with medical systematic review experience. We searched MEDLINE and EMBASE for articles written in English or French (1985–2010) on December 10, 2010, using terms related to neural tube defects (anencephaly or encephalocele or lipomeningocele or meningocele or myelomeningocele or neural tube defect or spina bifida) and epidemiology (incidence or prevalence or epidemiology). We also hand-searched the reference lists of review articles on the epidemiology of spina bifida and the references of included articles. We included studies reporting on data from 1985 onward because of the advances in neuroimaging and diagnosis since then.

Two study authors independently screened the abstracts and titles of all references in duplicate to find studies on the prevalence of spina bifida (excluding spina bifida occulta). We excluded data if they were not population-based. Two authors then reviewed the selected studies’ full text independently and in duplicate. Studies were eligible for the review if they met these criteria: (1) original research, (2) population-based (all cases in a defined geographic area or ascertainment from multiple hospitals or the only hospital in a defined area), and (3) reported an incidence or prevalence estimate or cases of spina bifida per population denominator. If there was disagreement about whether to include a study, the 2 authors would reach consensus through discussion or involve a third author if unable to reach consensus.

RESULTS

The search strategy yielded a total of 3336 citations: 1446 from MEDLINE and 1890 from EMBASE (Figure 1). After the initial screen, 738 articles met the criteria for full-text review, of which we excluded 600. Hand searching resulted in the inclusion of 41 additional articles. Characteristics of all 179 studies included in the systematic review are shown in Appendix C (available as a supplement to the online version of this article at http://www.ajph.org). From the 179 eligible studies overall, 123 studies included sufficient information to calculate period prevalence estimates between 1985 and 2010 and we thus included these in the meta-analyses.

Of the 179 included studies, all reported on the epidemiology of spina bifida.^{5–7,10,17,2–199} Of the studies, 92 reported on data from North America, 46 from Europe, 31 from Asia, 7 from Australia, 4 from South America, and 3 from Africa. (Some studies report on data from more than 1 continent.)

DISCUSSION

This study presents a comprehensive systematic review of the literature on the global prevalence of spina bifida. We performed meta-analyses to obtain estimates of spina bifida prevalence in subgroups of LBs, stillbirths, and TOPs in countries with or without mandatory folic acid fortification. Canada and the United States were the first countries to require mandatory fortification and multiple studies have documented a pre–post reduction in neural tube defects.⁶⁴ Currently, most studies in North America, South America, and Oceania have mandatory fortification; many African countries (e.g., Burkina Faso, Morocco)¹⁶ are implementing mandatory fortification. Fortification is uncommon in Asia and Europe. A higher prevalence of spina bifida was seen in countries without mandatory folic acid fortification policies in place at the time of data collection for the studies included in this systematic review. We found significant geographic variation in terms of spina bifida prevalence in studies including LBs, stillbirths, and TOPs; North American prevalence estimates were lower than all other continents. When LBs and stillbirths were reported together, North America had the lowest prevalence of spina bifida and Asia had the highest.

These regional differences reflect different regional policies with regard to folic acid fortification. It is also important to note the large number of prevalence estimates provided by ethnic subgroups in some American studies.^127,128 One study²⁰⁰ reported a period prevalence for Arizona in 1997 of 4.23 per 10 000 births in a non-Hispanic White sample whereas the corresponding prevalence estimate for the non-Hispanic Black or African samples was 0. Many African countries have recently instituted mandatory fortification and the impact of this in changing global rates of spina bifida will need to be assessed. Furthermore, access to prenatal screening, pregnancy termination, and differing folic acid fortification policies will also produce local and widespread variations in international comparisons such as this.

Policymakers are often challenged to make decisions with imperfect evidence of benefit or harm that have an impact on the health of the population.²⁰¹ Mandatory folic acid fortification for the prevention of neural tube defects is one such issue in which a global disparity exists. As of July 2012, 67 counties had fortified their wheat flour with folic acid (mandatory or voluntary programs), affecting approximately 2.2 billion people.¹⁶ As the neural tube closes early in gestation (day 28), public health campaigns aimed at encouraging pregnant women to use multivitamins containing folic acid often come too late, as many women do not realize that they are pregnant at this stage.⁹ An economic evaluation conducted in the United States almost 10 years after mandatory folic acid fortification estimated that the cost savings from this initiative ranged from $88 million to $142 million annually.²⁰² A systematic review of cost of illness studies similarly concluded that folic acid fortification was cost-effective in all studies.¹² This detailed review highlights the strong evidence in favor of mandatory and full coverage of grain fortification for prevention of spina bifida.^9,201

We identified several important sources of heterogeneity when extracting data for this systematic review. The first relates to the classification of infants with multiple malformations. Some sources only count the most serious birth defect whereas others count all defects. One review showed that including joint cases of anencephaly and spina bifida in prevalence estimates of spina bifida creates an inflation of the prevalence of spina bifida²⁰³; however, consistent international definitions of what explicitly should and should not be included in counts of spina bifida have not been established (e.g., studies that reported on spina bifida without anencephaly may underestimate the true prevalence).

The second source of heterogeneity involves the variability of methods of case ascertainment for spina bifida (e.g., birth defects surveillance systems, prospective clinical evaluation of all infants, retrospective chart review). The validity of each source of data has not been established and it is unknown what proportion of spina bifida cases are truly captured by each source. Different forms of ascertainment (e.g., birth certificates, surveillance systems) ascertain spina bifida in different ways with varying degrees of validity²⁰⁴; however, studies conducted to date suggest that almost all data sources suffer from some degree of underreporting^203,205 and US data suggest that prevalence estimates for spina bifida are robust across various sources of ascertainment, particularly those characterized by local and national collaboration.²⁰⁶ Reaching neonatal mortality targets set forth by Millennium Development Goal 4 (to reduce under-5 mortality by two thirds) requires local and national networks aimed at scaling up both active newborn health surveillance systems and newborn health programs.²⁰⁷

A third source of heterogeneity relates to the studies’ inclusion criteria (i.e., LBs-only compared with LBs, stillbirths, and TOPs). Differences remain in which populations are included in reported data, in that a sizable number of studies did not include data on TOPs. Although it is recognized that these data may be more difficult to obtain, this challenge should not negate the importance of this information in international comparisons. Differences in birth prevalence estimates may reflect differing genetic susceptibility, differential patterns of folic acid consumption, or different policies on TOP.²⁰³ In addition, because ultrasonography and prenatal biochemical screening are standard parts of prenatal care in many parts of the world, spina bifida is often detected prenatally. Because of differential patterns of pregnancy termination for fetuses with birth defects compared with nonmalformed fetuses, failure to account for pregnancy terminations likely introduces selection bias into prevalence estimates, underestimating observed associations.²⁰⁸

Spina bifida prevalence remains high in some regions despite policies promoting prenatal folic acid supplementation and fortification of the food supply. Health promotion strategies that encourage women of reproductive age to take folic acid supplements have failed.²⁰⁹ A recent study showed that the proportion of women taking their supplements in England decreased from 35% in 1999 to 2001 to 31% in 2011 to 2012.²⁰⁹ Fortification efforts should also address differences in the amount of folic acid that women receive from a fortified food supply. A recent review²¹⁰ showed that the target population of women of reproductive age who live in countries with mandatory folic acid fortification do not necessarily receive the daily recommended dose of 400 micrograms folic acid. In fact, the daily folic acid amount that is ingested in countries with national coverage programs can vary from 32.8 micrograms (Niger) to 736.7 micrograms (South Africa). Infrastructure that comprehensively addresses neonatal health surveillance, as well as micronutrient fortification, is vital for accelerating the reduction in neonatal mortality that we have observed, particularly in Asian and African regions.^21,211

This study is not without limitations. We did not include terms related to congenital anomalies and birth defects in the original search strategy as this study was specifically focused on spina bifida. This may have resulted in some studies being missed. However, because of the number of studies that we included and the fact that we hand-searched all of the reference lists of included articles, if any studies were not included, we do not anticipate them altering the pooled prevalence estimates. A second limitation concerns the differences between studies in defining the study population. There is no gold-standard diagnostic method for spina bifida and as such we may expect there to be a degree of underreporting in low- and middle-income regions where access to genetic testing and imaging could be lower. However, we found higher rates of spina bifida in low- and middle-income regions and we expect the risk of measurement bias to be negligible.

This systematic review provides global prevalence estimates of spina bifida across different regions, birth populations, and folic acid fortification policies. Spina bifida remains a significant source of worldwide infant mortality and morbidity despite decades-old evidence of the protective effects of dietary folic acid supplementation. Gross disparities exist across and between continents in terms of micronutrient food fortification legislation, food supply coverage, and per capita folic acid intake. Folic acid supplementation for spina bifida is not yet being used to its full potential and the worldwide implementation of mandatory folic acid fortification legislation is long overdue.