Assessing the association between natural food folate intake and blood folate concentrations: a systematic review and Bayesian meta-analysis of trials and observational studies

Abstract

Folate is found naturally in foods or as synthetic folic acid in dietary supplements and fortified foods. Adequate periconceptional folic acid intake can prevent neural tube defects. Folate intake impacts blood folate concentration; however, the dose-response between natural food folate and blood folate concentrations has not been well described. We estimated this association among healthy females. A systematic literature review identified studies (1 1992-3 2014) with both natural food folate intake alone and blood folate concentration among females aged 12-49 years. Bayesian methods were used to estimate regression model parameters describing the association between natural food folate intake and subsequent blood folate concentration. Seven controlled trials and 29 observational studies met the inclusion criteria. For the six studies using microbiologic assay (MA) included in the meta-analysis, we estimate that a 6% (95% Credible Interval (CrI): 4%, 9%) increase in red blood cell (RBC) folate concentration and a 7% (95% CrI: 1%, 12%) increase in serum/plasma folate concentration can occur for every 10% increase in natural food folate intake. Using modeled results, we estimate that a natural food folate intake of ≥ 450 μg dietary folate equivalents (DFE)/day could achieve the lower bound of an RBC folate concentration (~ 1050 nmol/L) associated with the lowest risk of a neural tube defect. Natural food folate intake affects blood folate concentration and adequate intakes could help women achieve a RBC folate concentration associated with a risk of 6 neural tube defects/10,000 live births.

Figure 1

Record management and selection flow diagram for the systematic review on natural food folate intake and blood folate concentrations. Wave 1 screening was a review of abstracts with a more broad focus. Wave 2 screening was a review of abstracts using the specific objectives from this study. Full text articles excluded (n = 183) is the sum of the 131 articles initially rejected after the full text review plus the 52 articles rejected after contacting authors. The final number of studies included in the qualitative synthesis (n = 36) is the sum of records accepted from full-text eligibility screening (n = 14), records eligible for inclusion through author contact (n = 19) and additional records identified through author contact (n =3).

Figure 2

Observed natural food folate intake [μg/day of DFE] versus RBC folate concentrations (nmol/L) among women of childbearing age from studies using a MA, no random study effect for intake, and posterior predicted distribution for future values of RBC folate concentration by natural food folate intake. Solid line represents the median value of the posterior predictive distribution; dotted lines represent the 95% CrI. The dots represent the standardized data points from the included studies. Included studies: [31,34,37,38,39,40] CrI: Credible interval; DFE: Dietary folate equivalent; MA: Microbiologic assay; RBC: Red blood cell.

Figure 3

Observed natural food folate intake (μg/day of DFE) versus serum/plasma folate concentrations (nmol/L) among women of childbearing age from studies using a MA, no random study effect for intake, and posterior predicted distribution for future values of serum/plasma folate concentration by natural food folate intake. Solid line represents the median value of the posterior predictive distribution; dotted red lines represent the 95% CrI. The dots represent the standardized data points from the included studies. Included studies: [31,34,37,38,39,40]. CrI: Credible interval; DFE: Dietary folate equivalent; MA: Microbiologic assay.

Figure 4

Observed natural food folate intake [μg/day of DFE] versus serum/plasma folate concentrations (nmol/L) among studies using a PBA, no random study effect for intake, and posterior predicted distribution for future values of serum/plasma folate concentration by natural food folate intake. Solid line represents the median value; dotted lines represent the 95% CrI. The dots represent the standardized data points from the included studies. Included studies: [41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57]. CrI: Credible interval; DFE: Dietary folate equivalent; PBA: Protein binding assay.