Purpose:

To assess the benefits and harms of folic acid supplementation in reproductive-age women for the prevention of neural tube defects (NTDs).

Data Sources:

Systematic review of studies from MEDLINE, the Cochrane Library, EMBASE unpublished literature, and trial registries through January 28, 2016; bibliographies from retrieved articles, outside experts, and reviewers.

Study Selection:

Two investigators independently selected studies using a priori inclusion and exclusion criteria. We included studies that focused on the use of folic acid supplementation (by itself or in multivitamin or prenatal supplement form) for the prevention of NTD-affected pregnancies in women of childbearing age. We limited the evaluation of benefits to NTDs.

We excluded poor-quality studies and studies of prepubertal girls, men or women without the potential for childbearing, and NTD recurrence prevention, and studies in developing countries.

Data Extraction:

One investigator extracted data and a second checked accuracy. Two reviewers independently rated the quality for all included studies, using predefined criteria.

Data Synthesis:

We included 24 studies: 12 on the effect of folic acid supplementation on NTDs, three on variation in the effect on NTDs by race/ethnicity, and eight on variation by dosage or timing. For harms, we focused on two recent systematic reviews on respiratory outcomes, which reported on several included studies. One systematic review also provided data on variation in harms by timing and dose. We paid particular attention in the analysis to variation in effect by study design and over time, particularly before and after food fortification in the United States (1998).

For the question of benefits of folic acid supplementation on NTDs, we found a single randomized, controlled trial (RCT), initiated in 1984 in Hungary, reporting a Peto odds ratio (OR) for NTDs of 0.131 (95% confidence interval [CI], 0.026 to 0.648; p=0.013). Two older cohort studies provided an OR of 0.11 (95% CI, 0.011 to 0.91) and 0.27 (95% CI, 0.11 to 0.63). Older case-control studies were also generally consistent with the evidence from the older RCT and cohort studies; odds ranged from 0.6 to 0.7 in three of four case-control studies. This evidence led to food fortification in the United States in 1998, after which no new trials or prospective cohort studies have been conducted. All newer evidence arises from case-control studies only. These newer studies, with inherently weaker designs, are consistent in not demonstrating a protective effect of folic acid supplements on NTDs, with odds ranging from 0.9 to 1.4 and CIs spanning the null.

Regarding variations in benefits by race/ethnicity, we found three eligible studies. One found no effect by race/ethnicity, a second found a higher but nonsignificant risk of NTDs with folic acid supplementation among Hispanic women (adjusted OR for consistent users compared with nonusers, 2.20 [95% CI, 0.98 to 4.92]), and a third found that the risk reduction was of smaller magnitude for Hispanic women compared with white or black women. These inconsistent results could have occurred by chance. Regarding variation in benefits, eight studies provided information. Of these, four studies provided information on dose, none on duration, and five on timing. We found no indication of a dose-response relationship in three of four studies. One study showed lower odds for daily versus less than daily use (OR, 0.57 [95% CI, 0.35 to 0.93]). Regarding timing, two older studies consistently showed no effect.

Two newer studies found no effect of timing for spina bifida, while one showed a protective effect with supplement use before pregnancy for anencephaly.

Regarding harms, one trial and one cohort study did not find evidence of statistically significant increased risk of twinning in women. The cohort study found that any increased risk of twinning was attenuated when the confounding effects of in vitro fertilization were accounted for. Three systematic reviews evaluated childhood asthma, wheezing, or allergy and found no consistent evidence of harm. One trial evaluated the risks of adverse events in women and found a higher risk of some events (weight gain, diarrhea, constipation) and not others (increased appetite, lack of appetite, exanthema, heartburn, and vertigo), but the event rate was very low and could have occurred by chance or as consequence of pregnancy. One systematic review did not find consistent evidence of an effect of folic acid supplementation on childhood asthma, wheezing, and allergy by timing or dose of intake.

Limitations:

We restricted interventions to folic acid supplementation and did not include interventions such as food fortification, counseling to increase dietary intake, or screening for NTDs. We found very limited information on differences in benefits and risks of folic acid supplementation by race/ethnicity, dose, and timing and no information on duration.

Regarding the overall quality of evidence, ethical considerations limit the use of RCTs to answer questions of efficacy. Observational studies carry limitations of case ascertainment and recall bias, and these two sources of bias can serve to reduce the observed effect of NTDs.

Conclusions:

Studies conducted before food fortification in the United States in 1998, with fewer design flaws, show that folic acid supplementation provides protection against NTDs. Newer studies, conducted after food fortification with folic acid, do not demonstrate this protective effect. These studies, however, have the potential for misclassification and recall bias, both of which can serve to attenuate the effect of folic acid supplementation on NTDs. Although mandatory food fortification in the United States has been accompanied by a decline in NTD prevalence, variations in intake continue to leave nearly a quarter of the U.S. population with suboptimal red blood cell folate concentrations, suggesting continued importance of folic acid supplement use.

Evidence of variations in effectiveness by race/ethnicity is inconsistent and could have occurred due to chance. We found no evidence of a dose-response effect, but studies had small numbers of cases for subanalyses. We did not find consistent evidence on timing of folic acid for benefits. We also did not find consistent evidence of harms, specifically twinning, respiratory outcomes, and other harms.