Introduction
Autism Spectrum Disorders (ASD) are characterized by representative impairments in social relatedness and verbal and non-verbal communication, combined with restrictive, repetitive patterns of behavior (APA,
2013). The prevalence of ASD had risen significantly from 13.4 per 1000 children in 2010 (CDC,
2014), 15.3 in 2012 (Christensen et al.,
2018), 17.0 in 2014 (Baio et al.,
2018) to 27.9 in 2016 (Xu et al.,
2019) in the USA. A recent meta-analysis indicated that the prevalence of ASD in China was 26.5/10,000, which demonstrated an increasing trend (Liu et al.,
2018) although it was significantly lower than the reported prevalence abroad. The impairments of ASD can severely impact learning and social functioning that may persist into adulthood (Cheuk et al.,
2011). Given that ASD is persistent disabling neural developmental disorders from early childhood, it had posed significant burdens on society and the economy (Baxter et al.,
2015; Leigh & Du,
2015).
The variation in ASD occurrence is mostly due to genetic factors (Bai et al.,
2019). However, environmental risk factors have been the focus of considerable attention in recent years (Schmidt et al.,
2019; Surén et al.,
2013). The hypothesis that the etiology of ASD is jointly mediated by genetic predispositions and environmental risk factors (Chaste & Leboyer,
2012; Gao et al.,
2020; Ijomone et al.,
2020; Kim et al.,
2019; Lyall et al.,
2017; Saxena et al.,
2020) has been widely accepted. Environmental factors may trigger predisposing hereditary high-risk gene modifications during fetal development (Oldenburg et al.,
2020; Saxena et al.,
2020). The pervasiveness of folate’s role in metabolism, nervous system function, and as a precursor of S-adenosyl-methionine (SAM), which is the folate donor for DNA methylation, suggests that there may be an association between maternal folate levels during pregnancy and ASD mediated by DNA methylation (Saxena et al.,
2020).
Folate is an essential water-soluble substance of the vitamin B family, which is present in natural foods, synthetic forms such as folic acid supplements, and food fortification products (Crider et al.,
2012; Mastroiacovo & Addis,
2006). As a crucial 1-carbon source, folate plays critical roles in cellular pathways such as DNA, RNA, and protein methylation DNA synthesis (Crider et al.,
2012). Before pregnancy, folic acid supplementation is recommended for all women of the reproductive age in order to decrease the risk of neural tube defects (MRC Vitamin Study Research Group,
1991; McGarel et al.,
2015; Valera-Gran et al.,
2014). Fortification of food products was also implemented in the United States and Canada (De Wals et al.,
2007; Pfeiffer et al.,
2019) but not in other countries, such as Sweden, Norway, Denmark, and Israel (Devilbiss et al.,
2017; Levine et al.,
2018; Surén et al.,
2013). Folic acid supplements during preconception may reduce the risk of autism and severe language delay in children (Roth et al.,
2011; Surén et al.,
2013). Abnormalities in folate metabolism may play a role in the occurrence of ASD (Frye et al.,
2017,
2018; Saha et al.,
2019; Shaik Mohammad et al.,
2016).
With the rising increase in prevalence and poor prognosis of ASD (Maenner et al.,
2020), it is imperative to enhance the prevention and treatment of this disease. Maternal use of folic acid supplements serves as a potentially crucial modifiable factor for fetal neurodevelopment. If maternal folic acid supplements can decrease the incidence of ASD, they can be used clinically in the prevention and treatment of ASD. Therefore, the potential of maternal use of folic acid supplements during pregnancy to prevent offspring ASD has been of increasing interest. Evidence derived from several studies on the association between maternal folic acid supplements and the prevention of ASD is inconclusive (Levine et al.,
2018; Li et al.,
2018; Surén et al.,
2013; Virk et al.,
2016). A case-cohort study from Israel, including 45,300 children, demonstrated that maternal exposure to folic acid supplements was associated with a lower risk of offspring's ASD. The information on folic acid supplement exposure was extracted from the Prescription Register, which significantly reduced recall bias (Levine et al.,
2018). Levine et al., also assessed the preventive effect of maternal exposure to folic acid supplements during pregnancy for offspring’s ASD with and without intellectual disability (ID) (Levine et al.,
2018). In the prospective Norwegian cohort study, folic acid intake from 4 weeks before to 8 weeks after conception was associated with a reduced risk of offspring's ASD (Surén et al.,
2013). However, other studies (Li et al.,
2018; Virk et al.,
2016) reported a null association between maternal folic acid supplementation and ASD risk. A limited number of meta-analysis studies reported controversial findings regarding the association between maternal folic acid supplements and offspring ASD (Guo et al.,
2019; Wang et al.,
2017). Several limitations are present in the previous systematic reviews as follows: (Guo et al.,
2019; Wang et al.,
2017): The included ASD cases partly diagnosed by screening scales (Jiang et al.,
2016; Sun et al.,
2016), the extracted data were from an inappropriate study (Al-Farsi et al.,
2013) without exploring the association between maternal folic acid intake and offspring ASD in the Wang’s review (Wang et al.,
2017); the inclusion of studies (Strøm et al.,
2018; Virk et al.,
2016) from duplicate sources, an unclear distinction regarding ORs, RRs and HRs (Levine et al.,
2018) following processing of the data, and misclassifications regarding the FA supplementation form (Surén et al.,
2013) in another review (Guo et al.,
2019); furthermore, high heterogeneity was noted without further exploration of meta-regression and sensitivity analysis in both reviews (Guo et al.,
2019; Wang et al.,
2017). Therefore, a comprehensive, critical, and updated review of the evidence is urgently required. In addition, crucial questions regarding the sensitive timing windows, suitable dosage, and appropriate supplement forms for maternal folic acid intake remain unexplored.
Therefore, a comprehensive systematic review and meta-analysis was conducted with the rigorous inclusion criteria to explore the association between maternal folic acid supplements and the risk of offspring's ASD. We also conducted a series of sensitive analysis and meta-regression to determine whether supplement timing, dose condition, supplementary mode, and folic acid food fortification moderated effect sizes. Given the public health and economic burden of autism, understanding the sensitive window and dose of maternal folic acid intake may efficiently and precisely improve prevention for the development of ASD in the offspring.
Discussion
The present study demonstrated that maternal folic acid supplements administered during the prenatal period were associated with 43% lower odds of offspring ASD compared to the subjects without maternal folic acid supplement exposure. Maternal folic acid supplements in early pregnancy may be the sensitive window to reduce the risk of offspring’s ASD. The minimum dosage of folic acid estimated to at least 400 μg daily may provide a protective effect on reducing the risk of offspring ASD. Folic acid intake and other nutrients or folic acid intake can reduce the risk of ASD in the offspring. The preventive effect of maternal folic acid supplements on offspring’s ASD during the prenatal period was explored in countries with or without folic acid food fortification. These findings provided evidence to support the critical role of maternal folic acid supplements in early pregnancy in reducing the risk of offspring ASD. The understanding of the sensitive supplement timing window and the suitable dosage of folic acid exposure may provide evidence for precise intervention and prevention.
Folic acid supplements consumed in early pregnancy or before pregnancy to early pregnancy could reduce the risk of offspring’s ASD (Fig.
3). These findings aid the identification of the optimal efficacy of folic acid in the protection against ASD development. Early pregnancy, notably in the first two months of pregnancy, is a critical period for the central nervous system development, mainly consisting of proliferation and migration of neural progenitor cells. Folate deficiency during early pregnancy may disrupt the proliferation and migration of neural progenitor cells by impairing the efficiency of DNA methylation (Crider et al.,
2012; Roffman,
2018) and resulting in brain abnormalities that may be associated with ASD (Choi,
2018; Cusick & Georgieff,
2012; Lieberman et al.,
2019; Roffman,
2018). Previous studies indicated that the period from deficiency to the recovery of serum folate levels should be estimated from several weeks to months. Early pregnancy folic acid supplements and periconceptional folic acid supplements should be emphasized and recommended to women of reproductive age. This hypothesis is supported by the sub-analyses (Fig.
3) and meta-regression analyses (Table
3). The preventive effect of maternal folic acid supplements on offspring ASD before pregnancy (Levine et al.,
2018; Li et al.,
2018) or during pregnancy (no detailed pregnancy period) (Levine et al.,
2018; Li et al.,
2018) against offspring’s ASD was not observed in the sub-analyses due to a lack of power due to lack of statistical power, which was attributed to the small number of studies. It is important to note that folic acid supplementation before pregnancy reduced the risk of neural tube defects (
1991; McGarel et al.,
2015; Valera-Gran et al.,
2014). Periconceptional folic acid supplements are widely considered to provide sufficient reserves for reducing the risk of neural tube defects and other neuropsychiatric risks (Roffman,
2018). The meta-regression analyses used in the present study verified further the protective effects of folic acid supplementation before pregnancy to early pregnancy on the offspring’s ASD. Although insufficient evidence was reported to support the hypothesis that maternal folic acid supplementation before pregnancy reduces the risk of offspring's ASD, the result should be considered with great caution due to the small number of studies. Future population-based studies are urgently required to explore the most sensitive time-window periods of maternal folic acid intake for the preventive effect on the offspring ASD.
Optimal folic acid dosage is essential for the pregnant woman health, safe preconception and for the development of the fetus. The meta-analysis focused on the exact dosage of maternal folic acid intake and indicated very low heterogeneity (Fig.
3). In addition, our subgroup analysis indicated that the risk of offspring’s ASD reduced from the onset when more than 400 μg folic acid was ingested daily. Maternal folic acid intake above the amount (≥ 400 μg) was already associated with lower ASD risk, as determined from the subgroup analysis. Recent studies reported that higher unmetabolized folic acid (UMFA) in cord blood at birth or maternal blood during pregnancy was associated with an increased risk of ASD (Egorova et al.,
2020; Raghavan et al.,
2018,
2020). The Boston Birth Cohort Study further observed that low and high folate concentrations in maternal plasma were associated with a higher risk of ASD, presenting a U-shaped risk curve (Raghavan et al.,
2018). During the prenatal period, high maternal folic acid exposure demonstrated several behavioral changes in an animal model, such as embryonic growth delay, memory impairment, anxiety-like behaviors, and methyl metabolism changes, which impacted ASD development (Bahous et al.,
2017). In addition, harmful effects on brain development were evident, which were due to higher folic acid intake. Therefore, it is necessary to explore the appropriate dosage of folic acid supplementation. Folic acid supplements are absorbed and can be rapidly converted to the 5,10-methylenetetrahydrofolate bioactive form (5,10-methlyTHF) by dihydrofolate reductase (DHFR) and methylenetetrahydrofolate reductase (MTHFR). DHFR in humans has sufficient capacity to efficiently metabolize the appropriate dosage of folic acid (upper intake level:1000 μg/day) (Bailey & Ayling,
2009; Obeid et al.,
2016). A randomized controlled trial also pointed out that low or undetectable concentrations of UMFA in serum of mothers and newborns were detected when folic acid intake was at a dose of 400 μg/day during pregnancy (Pentieva et al.,
2016). We found that maternal folic acid intake above the amount (≥ 400 μg) had already been associated with lower ASD risk from the subgroup analysis. We also observed that the dose above 500 or 800 μg/day did not offer additional benefit in the protection against ASD development compared to that noted at the dose higher than 400 μg/day. The safety of higher doses of maternal folic acid supplements had been reported (Krishnaveni et al.,
2014; Patel & Sobczyńska-Malefora,
2017; Raghavan et al.,
2018), and unintended negative consequences of excessive supplemental folic acid should be considered. Furthermore, increased maternal blood or fetal cord blood folate concentrations and an increased risk of ASD should be interpreted fully considering the folate metabolism abnormalities. When folate is present in the body, it is transported into the fetus and the brain primarily by the folate receptor alpha (FRα) mechanism (Desai et al.,
2017). Autoantibodies have been identified that can block folate binding to FRα (Desai et al.,
2017). These autoantibodies are highly prevalent in children with ASD and their mothers (Frye et al.,
2017). Abnormal function of the FRα transport mechanism can interfere with the transportation of folate into the brain and the fetus, potentially generating higher blood folate concentration levels, particularly if folic acid is being supplemented (Frye et al.,
2017). Folinic acid, a reduced form of folate, can be transported by the reduced folate carrier (RFC) when the FRα transport mechanism is disrupted byautoantibodies (Frye et al.,
2013). Folinic acid supplements have been proven to provide adequate folate to the brain and markedly improve the core associated ASD symptoms in children with ASD (Frye & Rossignol,
2014; Frye et al.,
2018). Folinic acid supplementation during the preconception and pregnancy period may be able to prevent offspring’s ASD. Future studies should not only focus on identifying the optimal dose of folic acid supplements to reduce the risk of offspring’s ASD but also on the difference in efficacy between folinic acid and folic supplements for the prevention of the development of offspring ASD during pregnancy.
Subgroup analysis indicated that either folic acid intake alone or a folic acid intake with other nutrients could reduce the risk of offspring’s ASD (Fig.
3). Other nutrients included multivitamins, nutrient specific vitamins, or minerals in the meta-analysis study. The present study demonstrated that folic acid intake only or folic acid intake with other nutrients could protect against offspring’s ASD development. The data did not suggest that folic acid intake with other nutrients had the advantage to reduce offspring's ASD than folic acid intake only. Adequate maternal nutritional status is critical for fetal brain development. In the rapid development period, the brain has heightened sensitivity to nutritional deficiency, which may predispose the fetus to postnatal neurodevelopmental disorders (Stephenson et al.,
2018). Additional population-based studies should rigorously investigate whether there is an independent or synergetic effect is evident between folic acid and other nutrients, which could drive the association with reduced ASD risk. In addition to folic acid (an oxidized form of folate), reduced folate forms, such as folinic acid, can be investigated with regard to the possible preventive effects on reducing offspring’s ASD. Unlike folic acid, folinic acid can enter the folate cycle without being converted by DHFR and MTHFR. Folinic acid can also pass the blood–brain barrier by RFC when the FRα is bound to FRAAs or is dysfunctional (Desai et al.,
2016; Frye et al.,
2013). In a randomized, double-blind placebo-controlled trial, folinic acid supplements were shown to improve verbal communication in children with ASD (Frye et al.,
2018). Whether folinic acid supplements during pregnancy prevent the offspring’s ASD has not been previously investigated. Future studies should explore whether folinic acid supplements may have an advantage over folic acid.
The present study explored whether countries and supplementary timings may be significant heterogeneity sources from meta-regression analyses (Table
3). Regardless of the folic acid food fortification policy implemented in countries, maternal folic acid supplements are associated with the reduced risk of children's ASD in the present study. The protective effect of maternal folic acid supplements was consistent in the US and Israeli populations, but not in the European or Chinese people. Several factors can be considered to account for the discrepancy in the results. Initially, the majority of European or Chinese studies did not report the consuming dose of FA intake, which may never reach the specific threshold of folic acid supplements to reduce the offspring’s ASD risk. Secondly, folic acid supplementary timing was a source of heterogeneity that suggested a sensitive period for maternal folic acid supplementation on the prevention of the offspring ASD.
To the best of our knowledge, this is the most comprehensive systematic review and meta-analysis regarding prenatal folic acid supplements and offspring’s autism spectrum disorders. In addition, it is the first to systematically account for supplementary timing, dose, supplemental folic acid mode, and folic acid food fortification. A series of sensitivity analyses were conducted in order to verify the stability and robustness of our results.
Several limitations should be mentioned with regard to the current study. Firstly, a recall bias was present since detailed information regarding maternal supplement intake before and during pregnancy was acquired using questionnaires. However, the majority of studies included in the meta-analysis focused on the analysis of continuous exposure over time rather than a single point time, which reflected a conservative estimate for ongoing folic acid supplement intake status (Schmidt et al.,
2012,
2017,
2019; Surén et al.,
2013; Virk et al.,
2016). Secondly, given the observational nature of the included studies, residual or unmeasured confounding factors are possible. Finally, a significant heterogeneity was observed in this meta-analysis. However, the heterogeneity was settled by subgroup analysis, meta-regression and a series of sensitivity analyses. The location of the study and the timing of folic acid intake were the significant sources of heterogeneity. The aforementioned variable factors should be considered to design future research studies exploring the association between maternal folic acid supplementation and ASD.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.