The association between maternal use of folic acid supplements during pregnancy and risk of autism spectrum disorders in children: a meta-analysis

We began our exploration into this topic through a study entitled “Environmental risk factors for autism: an evidence-based review of systematic reviews and meta-analyses” [1]. Within that paper, the authors analyzed the environmental risk factors for autism, including maternal use of folic acid supplements during pregnancy. Additionally, a study had been conducted to assess the association between folic acid and autism, which investigated that specific relationship more thoroughly [2]. However, these two previous studies did not combine their results using a meta-analytic approach. Therefore, the aim of this article is to establish a more definitive conclusion on the association between maternal use of folic acid supplements during pregnancy and autism spectrum disorder (ASD) risk in children.

The databases of PubMed, Web of Knowledge, and Wanfang Database were searched extensively to find eligible studies as recent as March 2017. The following search terms were used: [“folate” OR “folic acid” OR “vitamins” OR “diet”] AND [“Autism” OR “Autism spectrum disorder” OR “ASD”]. Articles were included if they presented data regarding the association between folic acid supplements during pregnancy and risk for ASD. Studies were excluded if they were written in languages other than English and Chinese, did not offer enough data, were reviews, editorials, comments, or if the studies were not conducted with humans as the test subjects. Two independent investigators searched and reviewed all identified studies, then extracted the data reported in Additional file 1: Table S1.

The DerSimonian and Laird random effects model was adopted to pool the study-specific relative risk (RR) with their 95% confidence intervals (CI) on the association between maternal use of folic acid supplements during pregnancy and ASD risk, which considers both within-study and between-study variation [3]. Heterogeneity across studies was checked by the chi-square test and I ² test, and I ² values of 0, 25, 50, and 75% represent no, low, moderate, and high heterogeneity, respectively [4, 5]. Sensitivity analysis was performed to assess whether the results could be affected markedly when one study was removed at a time [6]. Small study effect was assessed with visual inspection of Egger’s test and funnel plot [7]. Data was analyzed using STATA version 12.0 (Stata Corporation, College Station, TX, USA). P ≤ 0.05 (two-tailed) was accepted as statistically significant for computed effects.

The flowchart of this meta-analysis for exclusion/inclusion of studies is shown in Additional file 2: Figure S1. A total of 12 articles [8,9,10,11,12,13,14,15,16,17,18,19] with 16 studies comprising 4514 ASD cases were included in this report. Overall, maternal use of folic acid supplements during pregnancy could reduce the risk of ASD [RR = 0.771, 95% CI = 0.641–0.928, I ² = 59.7%, P _{heterogeneity} = 0.001, Fig. 1] as compared to those with no folic acid supplementation. Inverse relationships were found both in prospective studies [RR = 0.903, 95% CI = 0.786–0.998] and in case-control studies [RR = 0.435, 95% CI = 0.264–0.717]. The associations were also significant among Asian populations [RR = 0.669, 95% CI = 0.461–0.972], European populations [RR = 0.844, 95% CI = 0.683–0.993], and American populations [RR = 0.405, 95% CI = 0.165–0.994] (Fig. 1). The detailed subgroup analyses are showed in Table 1.

The forest plot of the association between maternal use of folic acid supplements during pregnancy compared with those no supplements, subgroup by ethnicity

Full size image

In the pooled result for all of the studies included, evidence of between-study heterogeneity was demonstrated. We therefore used univariate meta-regression to explore the reason for this high heterogeneity with the covariates of publication year, number of cases, study type, the time of folic acid supplement intake, the method of folic acid supplementation, and geographic locations. No significant finding was found contributing significantly to the between-study heterogeneity. Begg’s funnel plot (Additional file 3: Figure S2) and Egger’s test (P = 0.112) indicated that there was no significant publication bias. Sensitivity analysis showed that no single study had excessive influence on the pooled result while excluding one study at a time (Additional file 4: Figure S3).

One limitation of our report is that mothers were asked to recall a period several years before the interview when reporting on folic acid and supplement information. Given the elapsed time of the period of recall, inaccuracies would be expected. Another limitation is the significant heterogeneity observed in the result. The between-study heterogeneity was not resolved by meta-regression, and the results of subgroup analyses also demonstrated high heterogeneity. Therefore, other genetic and environment variables, as well as their possible interactions, may be potential contributors to the heterogeneity found.

In summary, this comprehensive meta-analysis suggested that maternal use of folic acid supplements during pregnancy could significantly reduce the risk of ASD in children as compared to those without folic acid supplementation. Significant associations were found in Asian, European, and American populations. Since some limitations are present in our study, future studies should aim to investigate the effect of the timing of folic acid supplement intake during the pregnancy, as well as the effects of varying dosages.