Following Smithells’ observations (1982) on the protective benefits of FA against NTDs and a decade-long controversy over his findings, two studies were decisive in mounting a public health effort to reduce the risk of NTDs.^8–10 The first was the Medical Research Council (1991) case-control study showing 72% reduction in NTDs among high risk pregnancies using 4.0 milligrams (mg) of FA in the periconceptional period.⁹ The second was the study by Czeizel and Dudas (1992) showing a similar reduction of the occurrence of NTDs using 0.8 mg of FA in the periconceptional period.¹⁰

Because South Carolina had been previously identified as a region with a high prevalence of NTDs, having approximately two times the national prevalence rate from 1973–1977,¹¹ the South Carolina NTD Surveillance and Prevention Program was initiated to capitalize on knowledge of the protective effect of FA.^12–14 The program followed recommendations from the Centers for Disease Control and Prevention (CDC) that all women of childbearing age should consume 0.4 mg of FA per day and that women with a previous NTD-affected pregnancy consume 4 mg of FA daily in the periconceptional period.^15,16 A boost came to the prevention effort in 1996 when the Food and Drug Administration mandated that enriched cereal grain flours be fortified with 0.14 mg FA per 100 grams milled flour.¹⁷ The rule, set to take place by January 1998, was accomplished by South Carolina flour producers by the summer of 1996. This report describes the identification and prevention of NTDs in South Carolina from 1992 through 2009.

Both active and passive surveillance methods were used to screen statewide for NTDs.^12,13 Active surveillance methods involved monitoring results of maternal α-fetoprotein laboratories, amniocentesis testing, and pregnancy ultrasonography programs. Passive surveillance methods included monthly reviews of medical reports from delivery hospitals and annual reviews of state birth and death certificates. All known live birth and fetal death cases from South Carolina were included. The surveillance period was from October 1 through September 30 of every year from 1992 through 2009. All infants that were conceived in South Carolina, including those delivered out-of-state, were included. Surveillance of NTDs was coordinated by the Greenwood Genetic Center from 1992 through June 2006 and thereafter by the South Carolina Department of Health and Environmental Control.

All cases were classified as spina bifida, anencephaly, or encephalocele. Co-occurring anencephaly and spina bifida (24 cases) were included in the anencephaly category; co-occurring spina bifida and encephalocele (four cases) were designated spina bifida. The two cases of aplasia cutis congenita were designated encephaloceles and the one case of multiple vertebral malformations was designated spina bifida. When possible, cases were examined by geneticists at the Greenwood Genetic Center and chromosome analyses were performed to search for coexisting anomalies and determine causation.

Efforts to decrease the occurrence of NTDs were targeted at health care professionals and the general public. Prevention alerts on FA and annual newsletters were distributed to obstetricians, family physicians, pediatricians, and health department clinics. A FA representative visited physician offices and health clinics to reinforce the FA message and provide patient literature and vitamins.

A public awareness campaign focused on women of childbearing age using billboards, television and radio announcements, newspaper releases, fact sheets, and brochures. The FA prevention message was included in continuing education courses for high school science teachers and in materials distributed at bridal fairs, sports events, and community festivals.

Once notified of an NTD-affected pregnancy, South Carolina Birth Defects Prevention Program personnel obtained permission from the primary physician to contact the mother. With such permission, a letter and a packet explaining recurrence prevention with FA were sent to the mother. The mother was contacted by telephone a week later to see if she was willing to participate in a behavioral risk survey and recurrence prevention program. Folic acid intake of 0.4 mg daily was recommended for women who were not actively attempting to get pregnant and 4.0 mg FA daily when actively trying to get pregnant. Enrolled women were continually monitored for FA supplement use and their reproductive plans. Women who were not actively trying to get pregnant were contacted by telephone every three months. Women who were pregnant or were actively trying to get pregnant were contacted every month and reminded of the recommended amount of FA intake.

During the initial five years of this project, the level of knowledge and use of FA supplementation was determined by interviews of a group of women (n=287) who had pregnancies which were not affected with an NTD. From 1997 to 2007, telephone surveys were made that included approximately 1,000 women each year who were in the childbearing years (ages 15 through 45 years). Among other questions, they were asked if they knew of the health benefits of FA, if they took FA supplements and, if so, how often.

Ultrasonography was the most frequent means of NTD detection in all years of surveillance, accounting for 55% of identification of cases in the first five years and for 78% in the last five years (Table II; available at www.jpeds.com). Maternal serum screening for alpha fetoprotein became a progressively less frequent source of initial case identification. Only 13% of NTD cases reached delivery without prior detection. Three sources – medical records, genetics clinics/autopsy records, and obstetric offices/clinics – accounted for 76% of case notifications to the surveillance program (Table III; available at www.jpeds.com).

At the outset of this study, spina bifida was the most common of the NTDs, accounting for 52% of the total number. Anencephaly was intermediate in prevalence and encephalocele least common, accounting for 35% and 13%, respectively.

Prevalence rates for spina bifida and anencephaly decreased significantly during the 17 year period with the prevalence rate for anencephaly exceeding the rate for spina bifida in some years (Table I). The prevalence rate for encephaloceles and for NTDs with associated anomalies did not decrease significantly. Among the 916 total NTDs identified, 728 (79%) were isolated defects.

A slight majority (106 of 188 – 56%) of NTDs with other structural anomalies constituted recognizable patterns of malformations (Table IV). Chromosome aberrations were found in 43 cases, with trisomy 18 (21 cases) being most common and triploidy (6 cases) being second most common. Other syndromes included amniotic bands (33 cases), Meckel syndrome (12 cases), OEIS (Omphalocele, Exstrophy of the bladder, Imperforate anus, and Sacral meningomyelocele; 7 cases), NTD-holoprosencephaly (6 cases), limb-body wall complex (2 cases), and Goldenhar syndrome (3 cases).

Eighty-two NTD cases with co-occurring anomalies did not appear to constitute recognizable syndromes or there was insufficient documentation to make a syndrome diagnosis (Table V; available at www.jpeds.com). The co-occurring anomalies ranged from mild and trivial (e.g. abnormal palmar creases, bifid scrotum) to severe and lethal (e.g. conjoined twinning, sirenomelia). Certain defects (e.g. diaphragmatic hernia, omphalocele) were placed in this category, although they might also have reasonably been classified as defects secondary to the NTDs.

Maternal diabetes mellitus has long been recognized as a risk factor for NTDs.⁴ We had information on the diabetes status of mothers in 704 pregnancies that resulted in infants born with an NTD. Twenty-eight of these (4.0%) had preexisting diabetes and 27 (3.8%) had gestational diabetes. The percentage of NTD-associated pregnancies with preexisting diabetes (4.0%) was fourfold the background rate of preexisting diabetes in South Carolina (0.9%). The percentage of NTD-associated pregnancies with gestational diabetes (3.8%) was not significantly different from the background rate of gestational diabetes (4.0%).¹⁸ After fortification, the rates of NTDs among infants of mothers with preexisting diabetes decreased from 3.5 to 2.7 per 100,000 live births and fetal deaths. The rate of NTDs among infants of mothers with gestational diabetes decreased from 6.4 to 1.4 per 100,000 after fortification.

NTDs (66 spina bifida, 55 anencephaly, and 19 encephalocele) occurred in pregnancies in which the mother reported use of FA during the periconceptional period (Table VI; available at www.jpeds.com). The dose of FA used by the mothers varied from 0.4 mg to 1.8 mg. A slightly higher, but non-significant (chi-square = 0.94, P=0.3321), percentage (23.6%) of these NTD pregnancies had associated anomalies, as compared with the percentage of NTD-associated anomalies (20.0%) born to mothers not using FA.

Pregnancies (n=484) occurred to women with a prior NTD. Among 418 pregnancies in which the mother used periconceptional FA supplements, there was one recurrence (0.2%), a recurrent anencephaly with a 13q deletion. Among 66 pregnancies in which the mother did not use FA supplements, there were four (6.1%) recurrences (a recurrent Meckel/encephalocele, a recurrent anencephaly, a recurrent encephalocele, and a situation where the first child had spina bifida and the second had craniorachischisis). These proportions are significantly different (p=0.0014) using a two-tailed Fisher exact test. Two of the recurrences (1.2%) occurred in 173 pregnancies prior to fortification, and the other three (1.0%) occurred in 311 post-fortification pregnancies. These proportions are not significantly different (p=1.0000) using a two-tailed Fisher exact test.

Based on the 1992 prevalence rates of 1.87 cases per 1,000 live births and fetal deaths, approximately 1,738 NTD-affected pregnancies would have been expected over the subsequent 16 years; however, there were only 885 NTD-affected pregnancies during this time period. This represents a reduction of 853 NTD cases prevented by FA and other unknown factors (Table I and Figure 2; available at www.jpeds.com). Since 1992–1993, there has been a decrease in the number of spontaneous abortions with NTDs (67% decrease), terminations of NTD-affected pregnancies (54% decrease), and live born infants with NTDs (54% decrease). Of the total 982 NTD-affected pregnancies over 17 years (Figure 2), 124 (12.6%) were lost to spontaneous abortion and fetal death, 430 (43.8%) were prenatally diagnosed and the pregnancies interrupted, and 428 (43.6%) were live born NTD cases.

The distribution among males and females with NTDs that had other associated anomalies was equal (87 males, 87 females, 14 with unknown sex), and the sex ratio of isolated NTDs showed significantly more females (47.5%) than males (39.7%, P-value = 0.003). There was a slightly higher prevalence rate of total NTDs among Hispanics (1.2 per 1,000 live births and fetal deaths) than any other racial group (non-Hispanic whites 1.1; non-Hispanic blacks 0.7). Although Hispanics had the highest rate of anencephaly, non-Hispanic whites had the highest rate of encephalocele as well as the highest rate of spina bifida. Hispanics had the highest rate of isolated NTDs, and all races had near equal rates of NTDs with other anomalies.

The latest rate for NTDs from the National Birth Defects Prevention Network covers the period 1998 (fourth quarter) to 2005 (fourth quarter). The “national” rate for spina bifida and anencephaly during this seven year period was 0.69 cases per 1000 live births and fetal deaths. ^19,20 The average rate for spina bifida and anencephaly in South Carolina during these same years was 0.69 cases per 1000 live births and fetal deaths. The rate for the most recent year (2009) for South Carolina was 0.65 cases per 1000 live births and fetal deaths, but the national rates for this period are not yet available.

With respect to occurrent NTDs, we also make these observations. First, the birth prevalence in our population has fallen to a level indistinguishable from the national NTD birth prevalence, suggesting that regional differences in NTD occurrence rates are either attributable to FA consumption patterns, or that the variability is confined to folate-responsive etiologic pathways. Second, we note that the reduction in the occurrence rates of spina bifida and anencephaly were more dramatically influenced by folate supplementation than the rate for encephalocele. Given the very high efficacy of FA in recurrence rates for subsequent pregnancies, we infer possible dosage response differences among different types of NTDs. Third, we can infer very little about racial differences in the rates for NTDs, and whether those differences reflect primarily genetic or environmental-cultural differences secondary to actual folate consumption patterns. Finally, we did observe a relatively dramatic reduction in NTD rates that coincided with the introduction of FA into wheat flour spread out over the interval of time from 1996 to 1998 when different sources of wheat flour were consumed by our population of mothers. This was preceded by a decline that mirrors increasing use of folate supplements that were likely attributable to the public health awareness campaign.

At the outset of the study, expectations were that FA supplementation might be approximately 70% efficacious in the prevention of recurrent NTDs.^8,9 There has been one recurrence of NTDs in 418 prospectively ascertained subsequent pregnancies (0.2%) to mothers with a prior conception or birth affected with an NTD who used FA in the periconceptional period of the subsequent pregnancy, the recurrence being an encephalocele associated with a 13q deletion. Thus, for isolated NTDs we observed 0.0% recurrence. The four recurrences among the 66 high risk pregnancies (6.1%) in which FA was not used emphasizes the importance of FA in recurrence prevention.^8,9 In the absence of folate supplementation, the recurrence rate of NTDs among families with a prior NTD conception or birth is generally stated to be 3–5%, with areas of higher prevalence having recurrence on the high end of that range. ²¹ For South Carolina, the recurrence risk without folate use would be at the high end, and thus we might have expected 21 or so cases of NTD among the 418 cases receiving folate supplements were folate not efficacious for prevention of NTDs, and approximately three cases among the 66 not receiving folate supplements. The number of pregnancies available for study is still fairly limited. However, we feel that our study and others²¹ have shown that recurrence prevention is quite effective. Because only approximately one-third of states have or have had NTD recurrence prevention projects,²² more public health agencies need to provide the funding and personnel to prevent the recurrence of NTDs in this very high risk population.

Because FA fortification appears to have been effective in decreasing the NTD rate among infants of mothers with gestational diabetes to the population rate, the NTD rate among infants of mothers with preexisting diabetes remained high after fortification. A greater effort to control preexisting diabetes may be necessary to further reduce the NTD rate in this high risk group. The probability that higher dose FA supplementation might be protective for infants of mothers with preexisting diabetes deserves further study.

Although several of the NTD cases with other anomalies may represent previously reported associations, there were no compelling reasons (known inheritance pattern, specificity of the association, confirmatory laboratory testing) to designate them as such. Among these previously reported associations are facial duplication-anencephaly association,²³ anophthalmia-NTD association,²⁴ coloboma-microphthalmia-clefting syndrome,²⁵ laterality syndrome,²⁶ schisis association,²⁷ anterior encephalocele-anophthalmia syndrome,²⁸ and others. Hunter has made an extensive listing of the numerous syndromes and associations of NTDs and other anomalies.² Specific recurrence risks are available for the NTD syndromes, but not for most of the entities with NTDs and other anomalies that do not constitute a specific syndrome.

Fifteen percent (15%) of NTDs occurred in pregnancies in which the mothers used FA in the periconceptional period. A slightly higher rate of syndromal NTDs was found in this group, suggesting that FA protection is more effective against isolated NTDs.

An increasing number of pregnant mothers receive birth defect screening through ultrasound and blood tests, resulting in an increasing number of NTDs that are diagnosed prenatally (Table III). Consequently, it seems important to point out that although termination of pregnancies affected with NTDs accounts for some fraction of the reduction of NTDs, it is the lower number of NTD-affected pregnancies conceived and not an increasing termination rate that is the driving force in the dramatic reductions we describe (Figure 2 and Table VII).