Background: Dietary reference values for folate intake vary widely across Europe. Methods: MEDLINE and Embase through November 2016 were searched for data on the association between folate intake and biomarkers (serum/plasma folate, red blood cell [RBC] folate, plasma homocysteine) from observational studies in healthy adults and elderly. The regression coefficient of biomarkers on intake (β) was extracted from each study, and the overall and stratified pooled β and SE (β) were obtained by random effects meta-analysis on a double log scale. These dose-response estimates may be used to derive folate intake reference values. Results: For every doubling in folate intake, the changes in serum/plasma folate, RBC folate and plasma homocysteine were +22, +21, and –16% respectively. The overall pooled regression coefficients were β = 0.29 (95% CI 0.21–0.37) for serum/plasma folate (26 estimates from 17 studies), β = 0.28 (95% CI 0.21–0.36) for RBC (13 estimates from 11 studies), and β = –0.21 (95% CI –0.31 to –0.11) for plasma homocysteine (10 estimates from 6 studies). Conclusion: These estimates along with those from randomized controlled trials can be used for underpinning dietary recommendations for folate in adults and elderly.

Adequate folate intake is necessary to achieve and maintain optimal health in all life stages. Folate deficiencies can result in many health problems, such as neural tube defects in developing embryos, and may lead to an increased risk of cardiovascular disease, anemia, cognitive impairment, and depression among adults and elderly [1-3]. Dietary reference values provide guidance on optimal dietary folate intake that should prevent deficiency and development of health problems.

The EURopean micronutrient RECommendations Aligned network (EURRECA; http://www.eurreca.org) has been developing approaches to derive and update micronutrient reference values, including folate. This process requires reliable data on the association between folate intake, status and health outcomes, which would allow estimating the intakes to achieve preset concentrations of relevant biomarkers. The EURRECA network has proposed 3 biomarkers of folate status: plasma or serum folate, red blood cell (RBC) folate, and plasma homocysteine (tHcy) [4]. A meta-analysis on dose-response relation between folic acid intake and plasma biomarkers from randomized controlled trials (RCTs) in adults and elderly has been recently investigated by EURRECA coworkers [3]. We now performed a systematic review that included data from observational studies in which folate intake represents intakes from its natural dietary matrix and from fortified foods and supplements, and in which populations are not confined to those strictly selected by the RCTs inclusion and exclusion criteria [5, 6]. This study fills a knowledge gap by systematically reviewing all available observational studies that investigated the relationship between folate intake and biomarkers of folate status in adults and elderly, followed by meta-analyses in order to model biomarkers of folate status as a function of dietary folate intake. Table 1 summarizes criteria used in the present systematic review.

Table 1.

PICOS criteria used to define the research question for the systematic review

PICOS criteria used to define the research question for the systematic review
PICOS criteria used to define the research question for the systematic review

Search Strategy

Within the EURRECA framework, systematic reviews were performed to explore the associations between intakes, status, and selected health outcomes for the public health priority micronutrients (iron, zinc, folate, vitamin B12, and iodine) [7-9]. This process followed a harmonized search strategy, aimed to collect the data published up to and including November 2016 using Medline and Embase search terms for (“study designs in -humans”) AND (folate or folic acid or vitamin B9) AND (intake OR status), -retrieved on Ovid Platform. Both indexing and text terms were used and languages included were restricted to those spoken in the EURRECA Network (English, Dutch, French, German, -Hungarian, Italian, Norwegian, Polish, Spanish, Greek, and -Serbian). The summary of the general search strategy developed for a search in EMBASE (on Ovid Platform) is shown in Table 2. Search strategies for MEDLINE (on Ovid Platform) were adapted based on this strategy. The reference lists of retrieved articles and published reviews were also checked for relevant studies. Where necessary, authors were contacted to provide missing data or clarify methods or results.

Table 2.

Summary of the EURRECA general search strategy including the search terms specific for folate, developed for a search in -EMBASE (Ovid)

Summary of the EURRECA general search strategy including the search terms specific for folate, developed for a search in -EMBASE (Ovid)
Summary of the EURRECA general search strategy including the search terms specific for folate, developed for a search in -EMBASE (Ovid)

Study Selection

For a study to be included, it had to meet all of the following inclusion criteria: designed as prospective cohort, nested case-control or cross-sectional study in apparently healthy adults and elderly; it had to report on both dietary folate intake and status following the EURRECA guidance on best practice micronutrient intake and status method [9]; folate intake was measured with a validated food frequency questionnaire (FFQ), a dietary history, or a 24-h recall (24HDR) or a food record/diary (FR); folate status was reported as plasma or serum folate, RBC folate, or total plasma homocysteine a non-specific marker of folate status.

Studies were excluded if they were retrospective cohort studies, or non-nested case-control studies, performed in non-healthy or populations other than adults and elderly, or were commentaries, reviews, or duplicate publications from the same study.

Of 7,249 identified articles in the wider search on folate intake, status and priority health outcomes in all populations, 7,049 were excluded following the screening of the title and abstract. The remaining 200 full-text articles were screened by 2 independent reviewers (R.N. and M.N.) and all discrepancies on study inclusion were discussed. This process resulted in 45 potentially relevant papers, out of which 23 were identified for the final analysis, in addition to 1 paper that was identified by following up on reference lists of the articles that were included into this review. Papers with incomplete data that could not be obtained from the authors were excluded. The flow diagram of the articles screened, assessed, and excluded at various stages is shown in Figure 1.

Fig. 1.

Flow diagram of the study selection process for the dose-response meta-analysis between folate intake and status biomarkers in adults and elderly. I, intake; S, status, referring to a biomarker of intake.

Fig. 1.

Flow diagram of the study selection process for the dose-response meta-analysis between folate intake and status biomarkers in adults and elderly. I, intake; S, status, referring to a biomarker of intake.

Close modal

Data Extraction

For each of the identified manuscripts, data were extracted into a standardized database. This was done by one (R.N.) and checked by another reviewer (M.N.).

Extracted data included population characteristics, mean, and SD of folate intake and dietary assessment method, concentrations of folate related biomarkers and analytical method used to measure folate status, the association and type of association between folate intake and folate-related biomarkers (Spearman rank correlation coefficient, Pearson correlation coefficient, linear regression coefficient), and information on any transformations applied to obtain the reported associations. Serum/plasma folate concentrations were converted to μmol/L when applicable.

In order to assess the quality of the included studies and the risk of bias, indicators of internal validity were collected during data extraction. Based on the indicators, 2 independent reviewers assessed the overall risk of bias and disagreements resolved by discussion. The criteria for judging these indicators were adapted from the Cochrane Library [10]. Table 3 presents the characteristics of the included studies [11-34].

Table 3.

General characteristics of the included observational studies in adults and elderly reporting the association between folate -intake and the biomarkers of folate status: plasma or serum folate, red blood cell folate, and plasma homocysteine

General characteristics of the included observational studies in adults and elderly reporting the association between folate -intake and the biomarkers of folate status: plasma or serum folate, red blood cell folate, and plasma homocysteine
General characteristics of the included observational studies in adults and elderly reporting the association between folate -intake and the biomarkers of folate status: plasma or serum folate, red blood cell folate, and plasma homocysteine

Data Synthesis

Each study provided at least one independent estimate; for some studies, 2 or 3 estimates were obtained because of stratification by gender [20, 25], by genotype [15, 17], by race [29, 32], or by diet type [23]. Where data on folate intake and status were available for males and females separately, but there was only one estimate of the association [13, 16, 25, 27]; the pooled estimates of mean folate intake and status were calculated and used in the analysis.

Statistical Analysis

The intake-status relationship was assumed to be linear with a coefficient between –1 and +1 on the loge–loge scale (natural logarithm of dependent and independent variable). Positive values would translate to a monotonic concave curve on the arithmetic scale that steeply increases at low intakes and less so at higher levels, which is a common shape in biology [6, 35].

Assuming this model, summary statistics from each study were transformed into estimates of the dose-response relationship: a regression coefficient (β) and its standard error (SE [β]) of micronutrient status on micronutrient intake. The individual estimates of the dose-response regression coefficients were combined into an overall pooled β and SE (β) by means of random effects meta-analysis, which incorporates the between-study variation using the method of DerSimonian and Laird [36] to obtain the weights required for the summary estimate. Residual heterogeneity between studies was evaluated using the I2 statistic. We investigated the intake-status relationship for each of the 3 biomarkers separately. In addition, we investigated whether mean age (continuous), dietary intake method (categories: [i] FFQ, and [ii] 24HDR or FR), and folate matrix (categories: [i] folate from the diet, [ii] folate from the diet and supplements), and analytical method used to measure folate status (categories: [i] microbiological assay, [ii] non microbiological assay) were variables that modified the association using meta-regression. Statistical transformations to obtain β and SE (β) were performed using GenStat version 13-SP2 (VSN International Ltd., http://www.vsni.co.uk/) and the meta-analysis was performed using STATA version 10.0 (College Station, TX, USA), with statistical significance defined as p < 0.05.

All studies that were included in this review were assessed for study quality by following the Cochrane Handbook. It was checked if these studies dealt with confounding factors adequately, whether assessment of exposure (intake or status) and funder were adequate, and if there were other potential threats to validity (online suppl. Table 1, see www.karger.com/doi/10.1159/000490003). Every study was assigned judgments such as “high,” “moderate,” or “low” risk for the purpose of study quality categorization. A consensus on the judgments was reached following discussion of each study. Thirteen studies [12, 15, 17, 19-25, 27, 30, 31] were judged at adequate risk of bias, 6 studies [14, 16, 18, 26, 29, 32, 34] were judged at a moderate risk of bias, and 4 studies was judged at a high risk of bias [11, 13, 28, 33].

Serum/Plasma Folate

We identified 26 estimates from 17 observational studies of folate intake and serum/plasma folate status that were eligible for our meta-analysis. Two papers reported results separately for 3 different genotypes, 1 paper reported results separately for 3 diet types, and 3 papers reported results for males and females separately. In total, 8,344 subjects were included with mean ages ranging from 21 to 65 years and sample sizes that ranged from 53 to 1,275 participants (Table 3). Most associations were obtained from cross-sectional analyses (n = 18), although some were derived from nested case-control studies (n = 6) and from a prospective cohort study (n = 1).

Combining these estimates in one meta-analysis yielded an overall pooled beta-coefficient of β = 0.29 (95% CI 0.21–0.37; I2 = 92.2%; Fig. 2) [11-27]. So, for studies in the range 152–662 µg/day, doubling of a folate intake resulted in an increase in serum/plasma folate by 2β or 22% (20.29 = 1.22). Thus, for example, an average person with a folate intake of 100 µg/day has a serum/plasma folate status concentration that is 22% higher than a person who has a folate intake of 50 µg/day, or an average person with a folate intake of 200 µg/day has a serum/plasma folate status concentration that is 22% higher than a person who has a folate intake of 100 µg/day.

Fig. 2.

Random effects meta-analysis of the association between dietary folate and serum/plasma folate in adults and elderly, stratified for the matrix of provided folate. The regression coefficients (β) represent the linear association between loge transformed folate intake and loge transformed serum/plasma folate status (lines represent the confidence intervals of each study). “Dietary folate” shows natural food folate expressed in μg/day in all studies; “Dietary folate including folic acid” shows natural food folate plus folic acid from supplements and fortified foods expressed in μg/day in all studies, except 3 in which it was expressed in DFE [14, 16, 22] (Table 3).

Fig. 2.

Random effects meta-analysis of the association between dietary folate and serum/plasma folate in adults and elderly, stratified for the matrix of provided folate. The regression coefficients (β) represent the linear association between loge transformed folate intake and loge transformed serum/plasma folate status (lines represent the confidence intervals of each study). “Dietary folate” shows natural food folate expressed in μg/day in all studies; “Dietary folate including folic acid” shows natural food folate plus folic acid from supplements and fortified foods expressed in μg/day in all studies, except 3 in which it was expressed in DFE [14, 16, 22] (Table 3).

Close modal

As the estimates showed substantial heterogeneity between the studies, the meta-analysis was subsequently stratified for folate matrix and dietary intake method. Stratification according to the matrix of provided folate (categories: [i] folate from the diet, [ii] folate from the diet and supplements, whether or not converted to dietary folate equivalents) yielded significantly different estimates of β = 0.33 (95% CI 0.22–0.44; n = 16) for dietary folate and β = 0.22 (95% CI 0.12–0.33; n = 10) for dietary folate plus supplements (Fig. 2). This difference was not statistically significant (p = 0.20) and the heterogeneity within the strata remained high (I2 = 92.3 and 86.6%).

Stratification for dietary intake method yielded different estimates for FFQs (β = 0.33, 95% CI 0.22–0.43; n = 17) and 24HDR or FR (β = 0.18, 95% CI 0.10–0.25; n = 9), although this may be attributed to chance, as the difference was not statistically significant (p = 0.19). The between-study heterogeneity (I2) remained high in the FFQ subgroup: it was 93.8% and moderate within 24HDR and FR – it was 38.7%.

Stratification for the analytical method used to measure serum/plasma folate (categories: [i] microbiological assay, [ii] non microbiological assay) yielded statistically different estimates of β = 0.51 (95% CI 0.37–0.64; n = 9) for microbiological assay and β = 0.18 (95% CI 0.14–0.22; n = 17) for non-microbiological assay (p < 0.0001). Studies that assessed serum/plasma folate response using the microbiological assay yielded somewhat higher heterogeneity (I2 = 87.3%) compared to studies that assessed folate response using non-microbiological assays (I2 = 49.2%).

The mean age of the study participants (continuous variable) was not a statistically significant determinant of the overall association (p = 0.52).

RBC Folate

We identified 11 observational studies of folate intake and RBC folate status that were eligible for our meta-analyses of which 4 also provided data on serum/plasma folate status. As 2 papers reported results by race separately, a total of 13 estimates were available for RBC folate, including a total of 3,997 subjects with mean age from 22 to 86 years, and sample sizes ranging from 53 to 760 subjects (Table 3). All studies were cross-sectional except 2 that dealt with prospective cohorts [32, 33], from which cross-sectional baseline data were used. Combining the 13 estimates in one meta-analysis yielded an overall pooled beta-coefficient β = 0.28 (95% CI 0.21–0.36; I2 = 88.1%; Fig. 3) [13, 14, 16, 26, 28-34]. So, for studies in the range from 177 to 592 µg/day for every doubling in folate intake, the increase in RBC folate concentration is 2β or 21% (20.28 = 1.21). Stratifying the analysis for the matrix of provided folate (categories: [i] folate from the diet, [ii] folate from the diet and supplements) yielded different estimates (β = 0.31, 95% CI 0.17–0.45, n = 6 and β = 0.27, 95% CI 0.17–0.36, n = 7 respectively), although they did not differ significantly and heterogeneity remained high (I2 = 75.7 and 92.3%, respectively; Fig. 3). Stratifying the analysis for the dietary intake method (categories: [i] FFQ and [ii] 24HDR or FR) yielded estimates of β = 0.16, 95% CI 0.12–0.19 (n = 4 estimates) for FFQ, and β = 0.33, 95% CI 0.25–0.41 (n = 9 estimates) for 24HDR or FR, and reduced the between-study heterogeneity in the FFQ subgroup, but not in the latter subgroup (I2 = 0 and 82.3% respectively). Stratification for analytical method used to measure RBC folate yielded similar estimates of β = 0.30 (95% CI 0.11–0.49 ; n = 3) for microbiological assay and β = 0.27 (95% CI 0.19–0.35; n = 10) for non-microbiological assay (p = 0.77). This subgroup analyses did not substantially reduce heterogeneity: I2 = 92.4 and 82.1% for microbiological and non-microbiological assay respectively. Each of these potential sources of heterogeneity (dietary matrix, dietary intake method, and analytical method used to measure RBC folate) was evaluated continuously and simultaneously in a meta-regression model. The results showed that none of the potential sources of heterogeneity, namely, the dietary intake method (p = 0.20), the matrix of provided folate (p = 0.95), analytical method used to measure RBC folate (p = 0.87), and the mean age of the study participants (continuous variable; p = 0.51) were statistically significant determinants of the overall beta.

Fig. 3.

Random effects meta-analysis of the association between dietary folate and red blood cell (RBC) folate in adults and elderly, stratified for the matrix of provided folate. The regression coefficients (β) represent the linear association between loge transformed folate intake and loge transformed RBC folate status (lines represent the confidence intervals of each study). “Dietary folate” shows natural food folate expressed in μg/day in all studies; “Dietary folate including folic acid” shows natural food folate plus folic acid from supplements and fortified foods expressed in μg/day in one study [26], whereas in others it was expressed in DFE [14, 16, 34] (Table 3).

Fig. 3.

Random effects meta-analysis of the association between dietary folate and red blood cell (RBC) folate in adults and elderly, stratified for the matrix of provided folate. The regression coefficients (β) represent the linear association between loge transformed folate intake and loge transformed RBC folate status (lines represent the confidence intervals of each study). “Dietary folate” shows natural food folate expressed in μg/day in all studies; “Dietary folate including folic acid” shows natural food folate plus folic acid from supplements and fortified foods expressed in μg/day in one study [26], whereas in others it was expressed in DFE [14, 16, 34] (Table 3).

Close modal

Plasma Homocysteine

We identified 6 observational studies of folate intake and plasma homocysteine status that were eligible for our meta-analyses; these studies also gave information on serum/plasma folate status. Two papers reported results separately for 3 different genotypes. In total, we had 10 estimates for plasma homocysteine including a total of 3,375 subjects with mean age from 28–65 years, and with sample sizes ranging from 81 to 938 subjects (Table 3). All studies were cross-sectional. Combining the 6 observational studies yielded an overall pooled beta of β = –0.21 (95% CI –0.31 to β = –0.11; I2 = 90.5%; Fig. 4) [13-15, 17, 25, 27]. Thus, for studies in the range from 201 to 331 µg/day, a doubling of folate intake goes together with a difference in plasma homocysteine concentration of 2β (2–0.26 = 0.84), which is –16%. Stratifying the analysis for the matrix of provided folate (categories: [i] folate from the diet, [ii] folate from the diet and supplements) yielded different estimates (β = –0.23, 95% CI –0.37 to –0.10 [n = 8 estimates] and –0.13, 95% CI –0.18 to –0.07 [n = 2 estimates] respectively Fig. 4). They did not differ significantly (p = 0.58) and the heterogeneity was reduced only in the latter subgroup (I2 = 0%), whereas in the former, it remained high (I2 = 91.7%).

Fig. 4.

Random effects meta-analysis of the association between dietary folate and plasma homocysteine folate in adults and elderly, stratified for the matrix of provided folate. The regression coefficients (β) represent the linear association between loge transformed folate intake and loge transformed plasma homocysteine folate status (lines represent the confidence intervals of each study). “Dietary folate” shows natural food folate expressed in μg/day in all studies; “Dietary folate including folic acid” shows natural food folate plus folic acid from supplements and fortified foods expressed in μg/day in one study [25], whereas in another [14] it was expressed in DFE (Table 3).

Fig. 4.

Random effects meta-analysis of the association between dietary folate and plasma homocysteine folate in adults and elderly, stratified for the matrix of provided folate. The regression coefficients (β) represent the linear association between loge transformed folate intake and loge transformed plasma homocysteine folate status (lines represent the confidence intervals of each study). “Dietary folate” shows natural food folate expressed in μg/day in all studies; “Dietary folate including folic acid” shows natural food folate plus folic acid from supplements and fortified foods expressed in μg/day in one study [25], whereas in another [14] it was expressed in DFE (Table 3).

Close modal

The mean age of the study participants (continuous variable) was a statistically significant determinant of the overall association (p = 0.004).

We did not perform stratified meta-analyses or meta-regression for the dietary intake method and total homocysteine because of the small number of studies in one of the strata. Also, stratified meta-analyses or meta-regression for the analytical method was not possible, as all studies used non-microbiological assay to determine plasma homocysteine levels.

The association between total folate intake (folic acid from supplements and/or fortified foods plus dietary folate) and its biomarkers in plasma/serum, RBCs, and plasma homocysteine levels in adults and elderly, stratified by folate matrix for plasma/serum, RBCs, and plasma homocysteine, and by dietary intake method and analytical method used to measure folate status biomarker for plasma/serum and RBCs is summarized in Table 4.

Table 4.

Associations between total folate supply (i.e., natural food folate plus folic acid from supplements and/or fortified foods) and folate in plasma/serum and in red blood cells, as well as on plasma homocysteine levels in adults and elderly, stratified by folate matrix, dietary intake method, and analytical method used to measure folate status

Associations between total folate supply (i.e., natural food folate plus folic acid from supplements and/or fortified foods) and folate in plasma/serum and in red blood cells, as well as on plasma homocysteine levels in adults and elderly, stratified by folate matrix, dietary intake method, and analytical method used to measure folate status
Associations between total folate supply (i.e., natural food folate plus folic acid from supplements and/or fortified foods) and folate in plasma/serum and in red blood cells, as well as on plasma homocysteine levels in adults and elderly, stratified by folate matrix, dietary intake method, and analytical method used to measure folate status

This systematic review gives a summary of available observational data on the association between folate intake and its biomarkers for adults and elderly. It uses best quality evidence from a large number of subjects with a broad range of folate intakes as from natural settings, and focuses on EURRECA proposed best practice biomarkers of folate status [5]. The main association found by pooling data from observational studies showed that total folate intake (i.e., folic acid plus dietary folate) in the range of 152–662 μg/day is significantly associated with folate status biomarkers: serum/plasma folate, RBC folate, and total plasma homocysteine concentrations in adults and elderly. For every doubling in folate intake, the changes in serum/plasma folate, RBC folate, and plasma homocysteine were +22, +21, and –16% respectively. This information, along with estimates from RCTs [3] can be used to decide what dose of folate or folic acid to recommend for optimal intake among adults and elderly.

For the interpretation of the results, it should be acknowledged that the studies included in this meta-analysis differed in terms of study populations and methods. For serum/plasma folate and RBC folate, the mean age was not a clear predictor of the association. Although the folate food matrix (for all 3 biomarkers), dietary assessment method (for RBC folate), and analytical method used to measure the folate status (for serum/plasma folate and RBC folate) affected the dose-response relationship, this explained only part of the large heterogeneity (I2 ranged from 88 to 92%) between the studies. Regarding the food matrix, we observed a stronger association for the estimate based on studies that included diet alone in comparison to those in which folate supply was from diet and supplements. This finding is in line with results from a similar systematic review of RCTs among women in childbearing age and in pregnancy and lactation [29]. Because synthetic folate is 1.7 times more bioavailable than naturally occurring folate [14, 37], total dietary folate should be expressed as DFE to account for this difference in bioavailability. However, for studies that assessed folate intake from the diet including the use of supplements, this conversion into DFE was done in only 5 out of 12 studies [14, 16, 22, 23, 32]. Also, some studies used 100–200 μg/day of folic acid as a conservative estimate of intake from supplements based on supplements available on the market, and added that amount without conversion to DFE to total dietary folate [18, 19, 25]. Finally, the proportion of regular and occasional supplement users varied between 4.8 and 26%, with average supplement dose up to 200 μg/day [11, 12, 17, 18, 22, 25, 27, 29]. This fact along with the lack of standardization in reporting total folate intake as DFE will have affected the estimates and contributed to heterogeneity of stratum-specific estimates. In addition, the stronger folate intake-status association resulting from natural folate in comparison to that from natural folate and folic acid could be linked to the property of naturally occurring folate: reportedly, it is as effective as or even more effective than folic acid in preserving folate status [35].

Studies included in these meta-analyses used a variety of analytical methods for measurement of folate status markers: 7 for serum/plasma, 5 for RBC folate, and 5 for plasma homocysteine. Because there were a small number of studies that applied the same analytical method per stratum for each biomarker, we were able to perform stratified meta-analyses or meta-regression only for the broad categories of microbiological assay and non-microbiological assay. This approach was already followed in a meta-analysis of folic acid intake-status relationship using the data from RCT [3]. We found statistically significant higher estimates for folate intake-status serum/plasma relationship when microbiological assay was used rather than non-microbiological assay, whereas the estimates for RBC folate did not statistically differ between the 2. Our finding that the type of analytical method had a considerable influence on the response, confirms the expert opinion that comparison of studies that used different assays requires caution, as different analytical methods are known to cause up to 30% difference in folate status results [38]. Indeed, future research should benefit from methodologically comparable data on folate intake and status.

We postulated that factors that explain heterogeneity in plasma/serum and RBC folate could be similar. Since plasma/serum and RBC folate yielded similar overall estimates, we combined the 22 estimates on serum/plasma folate and the 13 estimates on RBC folate in a meta-regression model, with the biomarker medium (serum/plasma or RBC folate) as an additional covariable. (Given that results from the same study are not statistically independent, we used studies [13, 14, 16, 23] to get the estimates for red blood folate only.) The biomarker medium in fact provided no statistically significant explanation (= 0.82). Our interpretation of this observation is that similar factors cause heterogeneity for both biomarkers.

Studies that were included in this review applied diverse dietary assessment methods too. It is another aspect that should be taken into consideration when examining the heterogeneity between studies, as it is known that dietary intake measurement errors differ considerably by dietary assessment method [39]. For example, in general, FFQs are designed to rank individuals rather than to assess their absolute intake levels. Also, a small number of replicates of 24HDR with no adjustment for intra-individual variability can give different estimates in comparison to multiple replicates. The FFQ was the most commonly used dietary instrument in this review; it was used to reflect folate intake for up to 3 months or 1 year. Stratified meta-analysis for the dietary intake method in serum/plasma and in RBC folate gave different estimates between the FFQ subgroup and 24HDR or FR subgroup: the association in FFQ subgroup was statistically and significantly weaker for RBC folate and statistically and not significantly stronger in plasma/serum folate when compared to 24HDR or FR subgroups per each biomarker. Observed differences in estimates indicate varying levels of bias associated with each measurement tool, for example, underreporting for 24 h recall, over- or underreporting of important folate-containing foods by the FFQ, inadequate folate values in food tables, different bioavailability of folate from different food products, the seasonality of data collection, or perhaps a combination of these conditions [40, 41]. For example, the nutrient intakes as from 24HDR is adjusted to reflect long-term exposure and, therefore, they may be more comparable to status assessed by RBC folate than by serum folate [42] as it implies from the results in this study. We hypothesized that, to some extent, the reason for the high heterogeneity in the meta-analysis of serum/plasma folate for the studies with FFQ might be genetic variation in the 677C→T mutation in the gene, which is known to significantly influence concentrations of folate status biomarkers at folate intake levels < 250 μg/day [15]. Because 2 studies provided the information on this gene mutation, we performed sensitivity analyses by removing the estimates with TT genotypes [15, 17]; however the heterogeneity remained the same (I2 = 94%). We also acknowledge that the estimation of folate intake requires reliable data on food composition. An important factor that hampers comparability of the data (between studies and across countries) and can be a source of errors is the lack of uniformity in food composition databases. This issue has been addressed in many past and ongoing projects worldwide. For instance, an important conclusion derived from the EFCOSUM project was that it is not possible at the present time to make existing food consumption data comparable at the nutrient level [45]. Therefore, for the purpose of this study, a problem of application of different food composition databases could not have been solved. Overall, when the pooled estimates from stratified subgroup analysis are analyzed in detail, folate matrix is the influencing factor on the overall beta for each folate-specific biomarker, in addition to the type of analytical method for serum/plasma folate (we were unable to do stratified analyses for dietary intake assessment and the type of analytical method at homocysteine because of small study numbers in each subgroup).

The relationship between folate intake and status in adults and the elderly has also been assessed in a meta-analysis of RCTs [3]. The association in RCTs was stronger than that in the observational studies in this paper, that is, for every doubling in folic acid intake, the increase in serum/plasma folate and RBC folate was 63 and 31% respectively. Another publication by Berti et al. [35] based on RCTs in women of reproductive age showed relationships for folate in RBCs and for total plasma homocysteine similar to those reported in the present study: 0.33 (95% CI 0.23–0.44) and –0.12 (95% CI –0.15 to –0.08), respectively, whereas the association for serum/plasma folate was stronger – 0.65 (95% CI 0.39–0.93) – than that shown by our data. One potential explanation for larger β in the RCTs could be higher bioavailabity of folic acid (85% or greater), which appears unaltered in the circulation and gives rise to serum/plasma folate in comparison to natural folates (25–50%) [43]. On the other hand, our estimates for serum/plasma as from subgroup analyses from microbiological assay are similar to those from RCTs. Assuming that microbiological assay is considered the gold standard for total folate measurement in serum [3], it may be that the overall pooled estimate for serum/plasma as from this paper is underestimated.

The observed mean folate intake in this review (266 μg/day) is in line with intakes indicated by individual observational studies that assessed folate intake in adults and elderly in Europe: the mean folate intake for adults and elderly in the European Nutrition and Health Report (25 countries) ranged from 130 to 370 μg/day in women and from 150 to 440 μg/day in men [44], whereas the results from the EPIC cohort (10 countries) reported 200–300 μg/day in women and 250–350 μg/day in men [45]. The levels of serum/plasma folate reported in a recent publication on micronutrient intake and status is Europe were 14–23 nmol/L [46], and in the publication by Duffy et al. [3], the levels were in the range 6–24 nmol/L and this is in line with the range observed in this review (5–31 nmol/L).

Thus, mean levels of intake and biomarkers are comparable to other studies, whereas their associations are heterogeneous. Although this heterogeneity might partially derive from methodological factors (standardization to DFE, dietary assessment method, analytical method used to determine level of status biomarkers), the overall estimates are consistent with those of other similar dose-response meta-analyses of trials (using synthetic folate) and observational studies among pregnant women. Possibly, to some extent, part of the between-study heterogeneity might reflect true differences between populations that average out in the meta-analysis.

This study shows that the quantified statistically significant relationships between folate intake and folate status biomarkers are similarly positive for serum/plasma and RBC folate and inverse for homocysteine. Because the ranges of dietary folate intake and serum/plasma folate shown in this review are similar to those reported in other studies conducted involving adults and elderly, the estimates shown by this review reflect folate intake and serum/plasma folate relationship in these population groups. The relative heterogeneity that we found between the studies indicates that for getting more reliable estimates, future research should take into consideration: (i) known differences in bioavailability between the natural forms and folic acid from supplements and fortified foods and (ii) variability that exists among different folate assay methodologies and across different laboratories [3]. Nevertheless, this study uses best quality available data and can be used to study the relationships between folate intake, status and health-related outcomes more in depth. Furthermore, reported estimates as from this paper together with those from RCTs can serve as a basis for further modeling toward determining dietary reference values for folate, such as the Average Nutrient Requirement (ANR), which vary greatly across countries (200–320 μg/day) [47-49]. The principle of that methodology has recently been developed within the EURRECA NoE and is readily available [6].

The authors have no relevant conflicts of interests to declare.

Funding and Sponsorship: This study has been carried out within the EURRECA Network of Excellence (http://www.eurreca.org), which is financially supported by the Commission of the -European Communities, specific Research, Technology and Development (RTD) Programme Quality of Life and Management of Living Resources, within the Sixth Framework Programme, project No. FP6 036196-2. This report does not necessarily reflect the Commission’s views or its future policy in this area. This work was in part funded by the Ministry of Education and Science of Republic Serbia, project III 41030.

R.N. conceptualized the paper, contributed to the literature searches and data extraction, performed the statistical analyses, and drafted the manuscript. P.V.V., D.R.-M., M.N. and A.G. supervised the work and assisted with analysis and editing of the manuscript. O.W.S., H.M.N., M.D., L.H., C.D. and J.M.S.R. contributed to the literature searches and data extraction. M.G., M.G., L..C.P.G.M.G., and P.V.V. were involved in development of the study hypothesis, editing the working versions of the manuscript and provided advice regarding -interpretation of the results. All authors participated in the writing or editing of the draft versions of the article and have read and approved of the final version of the submitted manuscript.

1.
Bermejo LM, Aparicio A, Rodríguez-Rodríguez E, López-Sobaler M, Andrés P, Ortega RM: Dietary strategies for improving folate status in institutionalized elderly persons. Br J Nutr 2009; 101: 1611–1615.
2.
Hoey L, McNulty H, Duffy ME, Hughes CF, Strain JJ: EURRECA-Estimating folate requirements for deriving dietary reference values. Crit Rev Food Sci Nutr 2013; 53: 1041–1050.
3.
Duffy ME, Hoey L, Hughes CF, Strain JJ, Rankin A, Souverein OW, Dullemeijer C, Collings R, Hooper L, McNulty H: Biomarker responses to folic acid intervention in healthy adults: a meta-analysis of randomized controlled trials. Am J Clin Nutr 2014; 99: 96–106.
4.
Fairweather-Tait SJ: Contribution made by biomarkers of status to an FP6 Network of Excellence, EURopean micronutrient RECommendations Aligned (EURRECA). Am J Clin Nutr 2011; 94: 651S–654S.
5.
Souverein OW, Dullemeijer C, van’t Veer P, van der Voet H: Transformations of summary statistics as input in meta-analysis for linear dose-response models on a logarithmic scale: a methodology developed within EURRECA. BMC Med Res Methodol 2012; 12: 57.
6.
Dullemeijer C, Souverein OW, Doets EL, van der Voet H, van Wijngaarden JP, de Boer WJ, Plada M, Dhonukshe-Rutten RA, In ’t Veld PH, Cavelaars AE, de Groot LC, van’t Veer P: Systematic review with dose-response meta-analyses between vitamin B-12 intake and -European Micronutrient Recommendations Aligned’s prioritized biomarkers of vitamin B-12 including randomized controlled trials and observational studies in adults and elderly. Am J Clin Nutr 2013; 97: 390–402.
7.
Moran VH, Stammers AL, Medina MW, Patel S, Dykes F, Souverein OW, Dullemeijer C, Pérez-Rodrigo C, Serra-Majem L, Nissensohn M, Lowe NM: The relationship between zinc intake and serum/plasma zinc concentration in children: a systematic review and dose-response meta-analysis. Nutrients 2012; 4: 841–858.
8.
Fekete K, Berti C, Trovato M, Lohner S, Dullemeijer C, Souverein OW, Cetin I, Decsi T: Effect of folate intake on health outcomes in pregnancy: a systematic review and meta-analysis on birth weight, placental weight and length of gestation. Nutr J 2012; 11–75.
9.
Dhonukshe-Rutten RA, Bouwman J, Brown KA, Cavelaars AE, Collings R, Grammatikaki E, de Groot LC, Gurinovic M, Harvey LJ, Hermoso M, Hurst R, Kremer B, Ngo J, Novakovic R, Raats MM, Rollin F, Serra-Majem L, Souverein OW, Timotijevic L, Van’t Veer P: EURRECA-evidence-based methodology for deriving micronutrient recommendations. Crit Rev Food Sci Nutr 2013; 53: 999–1040.
10.
HigginsJP, Green S: Cochrane Handbook for Systematic Reviews for Interventions, version 5.1.0 (updated March 2011); The Cochrane Collaboration, 2011. http://handbook-5–1.cochrane.org/(accessedonFebruary1, 2018).
11.
Arnaud J, Fleites-Mestre P, Chassagne M, Verdura T, Garcia Garcia I, Hernandez-Fernandez T, Gautier H, Favier A, Pérez-Cristià R, Barnouin J: Vitamin B intake and status in healthy Havanan men, 2 years after the Cuban neuropathy epidemic. Br J Nutr 2001; 85: 741–748.
12.
Ashfield-Watt PA, Whiting JM, Clark ZE, Moat SJ, Newcombe RG, Burr ML, McDowell IF: A comparison of the effect of advice to eat either “5-a-day” fruit and vegetables or folic acid-fortified foods on plasma folate and homocysteine. Eur J Clin Nutr 2003; 57: 316–323.
13.
Chew SC, Khor GL, Loh SP: Association between dietary folate intake and blood status of folate and homocysteine in Malaysian adults. J Nutr Sci Vitaminol 2011; 57: 150–155.
14.
Colić Barić I, Satalić Z, Keser I, Cecić I, Sucić M: Validation of the folate food frequency questionnaire with serum and erythrocyte folate and plasma homocysteine. Int J Food Sci Nutr 2009; 60(suppl 5): 10–8.
15.
de Bree A, Verschuren WM, Bjørke-Monsen AL, van der Put NM, Heil SG, Trijbels FJ, Blom HJ: Effect of the methylenetetrahydrofolatereductase 677C–>T mutation on the relations among folate intake and plasma folate and homocysteine concentrations in a general population sample. Am J ClinNutr 2003; 77: 687–693.
16.
Drogan D, Klipstein-Grobusch K, Wans S, Luley C, Boeing H, Dierkes J: Plasma folate as marker of folate status in epidemiological studies: the European Investigation into Cancer and Nutrition (EPIC)-Potsdam study. Br J Nutr 2004; 92: 489–496.
17.
Fukuda N, Hamajima N, Wakai K, Suzuki K: A cross-sectional study to find out the relationship of methylenetetrahydrofolate reductase (MTHFR) C677T genotype with plasma levels of folate and total homocysteine by daily folate intake in Japanese. J Nutr Sci Vitaminol (Tokyo) 2014; 60: 231–238.
18.
Glynn SA, Albanes D, Pietinen P, Brown CC, Rautalahti M, Tangrea JA, Gunter EW, Barrett MJ, Virtamo J, Taylor PR: Colorectal cancer and folate status: a nested case-control study among male smokers. Cancer Epidemiol Biomarkers Prev 1996; 5: 487–494.
19.
Johansson I, Van Guelpen B, Hultdin J, Johansson M, Hallmans G, Stattin P: Validity of food frequency questionnaire estimated intakes of folate and other B vitamins in a region without folic acid fortification. Eur J ClinNutr 2010; 64: 905–913.
20.
Melse-Boonstra A, de Bree A, Verhoef P, Bjørke-Monsen AL, Verschuren WM: Dietary monoglutamate and polyglutamatefolate are associated with plasma folate concentrations in Dutch men and women aged 20–65 years. J Nutr 2002; 132: 1307–1312.
21.
Sedjo RL, Inserra P, Abrahamsen M, Harris RB, Roe DJ, Baldwin S, Giuliano AR: Human papillomavirus persistence and nutrients involved in the methylation pathway among a cohort of young women. Cancer Epidemiol Biomarkers Prev 2002; 11: 353–359.
22.
Shuaibi AM, Sevenhuysen GP, House JD: Validation of a food choice map with a 3-day food record and serum values to assess folate and vitamin B-12 intake in college-aged women. J Am Diet Assoc 2008; 108: 2041–2050.
23.
Schüpbach R, Wegmüller R, Berguerand C, Bui M, Herter-Aeberli I: Micronutrient status and intake in omnivores, vegetarians and vegans in Switzerland. Eur J Nutr 2017; 56: 283–293.
24.
Sivaprasad M, Shalini T, Balakrishna N, -Sudarshan M, Lopamudra P, Suryanarayana P, Arlappa N, Ravikumar BP, Radhika MS, Reddy GB: Status of vitamin B12 and folate among the Urban adult population in South India. Ann Nutr Metab 2016; 68: 94–102.
25.
Van Guelpen B, Hultdin J, Johansson I, Witthöft C, Weinehall L, Eliasson M, Hallmans G, Palmqvist R, Jansson JH, Winkvist A: Plasma folate and total homocysteine levels are associated with the risk of myocardial infarction, independently of each other and of renal function. J Intern Med 2009; 266: 182–195.
26.
Verkleij-Hagoort AC, de Vries JH, Stegers MP, Lindemans J, Ursem NT, Steegers-Theunissen RP: Validation of the assessment of folate and vitamin B12 intake in women of reproductive age: the method of triads. Eur J Clin Nutr 2007; 61: 610–615.
27.
Weinstein SJ, Albanes D, Selhub J, Graubard B, Lim U, Taylor PR, Virtamo J, Stolzenberg-Solomon R: One-carbon metabolism biomarkers and risk of colon and rectal cancers. Cancer Epidemiol Biomarkers Prev 2008; 17: 3233–3240.
28.
Fayet F, Flood V, Petocz P, Samman S: Relative and biomarker-based validity of a food frequency questionnaire that measures the intakes of vitamin B(12), folate, iron, and zinc in young women. Nutr Res 2011; 31: 14–20.
29.
Fraser GE, Jaceldo-Siegl K, Henning SM, Fan J, Knutsen SF, Haddad EH, Sabaté J, Beeson WL, Bennett H: Biomarkers of dietary intake are correlated with corresponding measures from repeated dietary recalls and food-frequency questionnaires in the adventist health study-2. J Nutr 2016; 146: 586–594.
30.
Hoey L, McNulty H, Askin N, Dunne A, Ward M, Pentieva K, Strain J, Molloy AM, Flynn CA, Scott JM: Effect of a voluntary food fortification policy on folate, related B vitamin status, and homocysteine in healthy adults. Am J Clin Nutr 2007; 86: 1405–1413.
31.
Hopkins SM, Gibney MJ, Nugent AP, McNulty H, Molloy AM, Scott JM, Flynn A, Strain JJ, Ward M, Walton J, McNulty BA: Impact of voluntary fortification and supplement use on dietary intakes and biomarker status of folate and vitamin B-12 in Irish adults. Am J Clin Nutr 2015; 101: 1163–1172.
32.
Knutsen SF, Fraser GE, Linsted KD, Beeson WL, Shavlik DJ: Comparing biological measurements of vitamin C, folate, alpha-tocopherol and carotene with 24-hour dietary recall information in nonhispanic blacks and whites. Ann Epidemiol 2001; 11: 406–416.
33.
Mendonça N, Mathers JC, Adamson AJ, Martin-Ruiz C, Seal CJ, Jagger C, Hill TR: Intakes of folate and vitamin B12 and biomarkers of status in the very old: the newcastle 85+ study. Nutrients 2016; 8:pii:E604.
34.
Owens JE, Holstege DM, Clifford AJ: Comparison of two dietary folate intake instruments and their validation by RBC folate. J Agric Food Chem 2007; 2: 3737–3740.
35.
Berti C, Fekete K, Dullemeijer C, Trovato M, Souverein OW, Cavelaars A, Dhonukshe-Rutten R, Massari M, Decsi T, Van’t Veer P, Cetin I: Folate intake and markers of folate status in women of reproductive age, pregnant and lactating women: a meta-analysis. J NutrMetab 2012; 2012: 470656.
36.
DerSimonian R, Laird N: Meta-analysis in clinical trials. Controlled Clin Trials 1986; 7: 177–188.
37.
Suitor CW, Bailey LB: Dietary folate equivalents: Interpretation and application. J Am Diet Assoc 2000; 100: 88–94.
38.
Öhrvik V, Lemming EW, Nälsén C, Becker W, Ridefelt P, Lindroos AK: Dietary intake and biomarker status of folate in Swedish adults. Eur J Nutr 2016,Epub ahead of print.
39.
Dodd KW, Guenther PM, Freedman LS, et al: Statistical methods for estimating usual intake of nutrients and foods: a review of the theory. J Am Diet Assoc 2006; 106: 1640–1650.
40.
Brevik A, Vollset SE, Tell GS, Refsum H, Ueland PM, Loeken EB, Drevon CA, Andersen LF: Plasma concentration of folate as a biomarker for the intake of fruit and vegetables: the Hordaland Homocysteine Study. Am J Clin Nutr 2005; 81: 434–439.
41.
Marchetta CM, Devine OJ, Crider KS, Tsang BL, Cordero AM, Qi YP, Guo J, Berry RJ, Rosenthal J, Mulinare J, Mersereau P, Hamner HC: Assessing the association between natural food folate intake and blood folate concentrations: a systematic review and Bayesian meta-analysis of trials and observational studies. Nutrients 2015; 7: 2663–2686.
42.
Pfeiffer CM, Schleicher RL, Johnson CL, Coates PM: Assessing vitamin status in large population surveys by measuring biomarkers and dietary intake – two case studies: folate and vitamin D. Food Nutr Res 2012; 56.
43.
Vitamin and Mineral Requirements in Human Nutrition. Second Edition of Report of A Joint FAO/WHO Expert Consultation, Bangkok, Thailand. World Health Organization (WHO), Food and Agriculture Organization of the United Nations (FAO). 2004. (ISBN 92 4 1546123).
44.
Elmadfa I, Meyer A, Nowak V, Hasenegger V, Putz P, Verstraeten R, Remaut-DeWinter AM, Kolsteren P, Dostálová J, Dlouhý P, Trolle E, Fagt S, Biltoft-Jensen A, Mathiessen J, Velsing Groth M, Kambek L, Gluskova N, Voutilainen N, Erkkilä A, Vernay M, Krems C, Strassburg A, Vasquez-Caicedo AL, Urban C, Naska A, Efstathopoulou E, Oikonomou E, Tsiotas K, Bountziouka V, Benetou V, Trichopoulou A, Zajkás G, Kovács V, Martos E, Heavey P, Kelleher C, Kennedy J, Turrini A, Selga G, Sauka M, Petkeviciene J, Klumbiene J, Holm Totland T, Andersen LF, Halicka E, Rejman K, Kowrygo B, Rodrigues S, Pinhão S, Ferreira LS, Lopes C, Ramos E, Vaz Almeida MD, Vlad M, Simcic M, Podgrajsek K, Serra Majem L, Román Viñas B, Ngo J, Ribas Barba L, Becker V, Fransen H, Van Rossum C, Ocké M, Margetts B: European nutrition and health report 2009. Forum Nutr 2009; 62: 1–405.
45.
Park JY, Nicolas G, Freisling H, Biessy C, Scalbert A, Romieu I, Chajès V, Chuang SC, Ericson U, Wallström P, Ros MM, Peeters PH, Mattiello A, Palli D, María Huerta J, Amiano P, Halkjær J, Dahm CC, Trichopoulou A, Orfanos P, Teucher B, Feller S, Skeie G, Engeset D, Boutron-Ruault MC, Clavel-Chapelon F, Crowe F, Khaw KT, Vineis P, Slimani N: Comparison of standardised dietary folate intake across ten countries participating in the European Prospective Investigation into Cancer and Nutrition. Br J Nutr 2012; 108: 552–569.
46.
Novaković R, Cavelaars AE, Bekkering GE, Roman-Viñas B, et al: Micronutrients intake and status in Central and Eastern Europe as compared to other European countries, results from the EURRECA network. Public Health Nutr 2013; 16: 824–840.
47.
Nordic Nutrition Recommendations 2004, ed 4. Copenhagen, Nordic Council of Ministers, Nord, 2004, p 13.
48.
Dietary reference intakes for thiamin, riboflavin, niacin, vitamin B6, folate, vitamin B12, pantothenic acid, biotin, and choline.  -Washington, Institute of Medicine (US) Standing Committee on the Scientific Evaluation of Dietary Reference Intakes and its Panel on Folate, Other B Vitamins, and Choline. Washington (DC): National Academies Press (US); 1998. ISBN-10: 0–309–06411–2ISBN-13: 978–0-309-06411-8.
49.
Requirements of Vitamin A, Iron, Folate and Vitamin B12. FAO/WHO expert consultation. World Health Organization (WHO), Food and Agriculture Organization of the United Nations (FAO), 1998, pp 51–61.
Copyright / Drug Dosage / Disclaimer
Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.