How well do blood folate concentrations predict dietary folate intakes in a sample of Canadian lactating women exposed to high levels of folate? An observational study

Folic acid supplementation during the periconceptional period is effective in reducing the risk of neural tube defects (NTDs) [1–5]. However, the proportion of Canadian women that report consuming a folic acid-containing supplement during the periconceptional period is low ranging between 17 – 28% [6–9]. Similar rates are documented even among women who are knowledgeable about the health benefits of consuming folic acid during this life-stage [8]. As a result, mandatory folic acid fortification of white flour and select cereal grain products was implemented in 1998 in Canada with the intention to increase the dietary folate intakes of reproducing women by 80 to 100 μg/d [10]. Indeed folic acid fortification has resulted in a dramatic increase in blood folate concentrations in North America, and a significant reduction in NTD-affected pregnancies (19 – 78%) [7, 11–14].

In response to the improved blood folate concentrations of reproductive age women, many health care professionals are requesting direction on how aggressively they should promote folic acid supplementation among patients with high blood folate concentrations during the periconceptional period. The research literature shows many Canadian women post-fortification have red blood cell (RBC) folate concentrations approaching or exceeding 906 nmol/L [7, 15], a value in early pregnancy shown to be an indicator of a reduced risk for NTDs [16–18]. For example, we previously reported median RBC folate concentrations ≥ 3000 nmol/L during the 3^rd trimester of pregnancy in a sample of well-educated Canadian women [15].

It must be acknowledged that the body of evidence supporting a relationship between folic acid supplementation during the periconceptional period and the risk of an NTD-affected pregnancy is considerably more robust than that of the association between blood folate concentrations and NTD risk [1–5, 19]. Nonetheless, it is hard to argue against the logic that RBC folate concentrations, a measure of tissue folate stores, are unlikely to be elevated unless dietary folate intakes are similarly high [20]. For example, in controlled feeding studies preceded by a folate depletion phase of 2–7 weeks, serum and RBC folate concentrations of reproductive age women reflect the folate content of their diet (folate depletion, 400, 800 μg/day dietary folate equivalents [DFE]) [21–25]. Pre-folic acid fortification of the food supply, we reported a weak association between RBC folate concentrations and dietary folate intakes in young women not consuming a folic acid-containing supplement (r = 0.23, P < 0.05) [26]. This association improved when young women consuming supplemental folic acid were included in the analyses (r = 0.50, P < 0.01), suggesting that as the proportion of folate intake from synthetic folic acid increases, so might the predictive value of blood folate measures in determining folate intake. We speculated that folic acid added as a fortificant to food may likewise strengthen this association given its stability, and improved precision and accuracy of measurement, compared to endogenous folates which are normally trapped within cellular matrices, and need to be converted to a microbiologically assayable form. Second, inclusion of women known to consume supplements and fortified foods is likely to increase the range of exposure to folate which may also strengthen the association between blood folate concentration and folate intake. In contrast, if the relationship between blood folate concentration and folate intake is no longer linear at higher levels of exposure or the quality of the food composition tables has deteriorated given the variability in mandated versus actual levels of folic acid fortification, the association between blood folate concentration and dietary folate intake may not have improved.

Lactation is an opportune time to investigate the predictive value of blood folate concentrations in determining dietary folate intake as (1) women are clearly at a reproductive stage of their life cycle; (2) requirements for folate are elevated [27]; (3) vitamin supplement use is common. Therefore, the objective of this study was to examine whether, or not, blood folate concentrations, and dietary and synthetic folic acid intakes are predictive of each other in a sample of affluent, well-educated lactating women exposed to high levels of folate in the form of a supplement and fortificant.

Results from the present study suggest that blood folate concentrations, and folate intake determined by food composition tables, are not reliable predictors of each other in a sample of Canadian lactating women exposed to high levels of folate. In fact, plasma and RBC folate concentrations predicted the correct quartile of folate intake of lactating women in our study less than half of the time. Arguably the best evidence supporting a relationship between folate and NTD-prevention comes from randomized prevention trials in which women were assigned to receive supplemental folic acid (± other vitamins and minerals) during the periconceptional period compared to those that were not [1, 3–5]. Given the uncertainty of the mechanism(s) of how folate reduces the risk of NTDs, many health care professionals ask how well RBC folate concentrations predict synthetic folic acid intake in order to inform their decision on whether to recommend folic acid supplementation to women with very high blood folate concentrations. While it is conceivable that post-folic acid fortification of the food supply RBC folate concentration may be a better predictor of NTD risk than estimated folate intake, and that folic supplementation is no longer necessary for women with high RBC folate concentrations, randomized controlled trials to confirming this are unlikely to be conducted due to ethical considerations. Other experimental approaches will need to be considered. These known uncertainties together with our data illustrating that blood folate concentrations are not very predictive of folate intake suggest it would be prudent for women even with high blood folate concentrations to consume a folic acid supplement during the periconceptional period for NTD prevention.

Contrary to our original hypothesis, the strength of the association between dietary folate intake, determined by food composition tables, and blood folate concentrations has not improved post-folic acid fortification of the food supply. In the present study, statistically significant correlations between dietary folate intake, as determined by 3-d weighed food records, and RBC folate, an indicator of NTD risk, at 4- (r = 0.32, P < 0.05) and 16-wk (r = 0.36, P < 0.01) postpartum were found. The strength of these associations are very similar to those recently reported between dietary folate intake, as assessed by an abbreviated folate-targeted food/supplement screening tool, and RBC folate concentration, in a sample of women from California (r = 0.32, P = 0.0001) [35]. Neither the correlations reported in the present study nor those in the California report are stronger than those we previously reported pre-fortification in a group of young women from Southern Ontario (r = 50, P < 0.01) [26].

There are at least two likely explanations for our observations. First, the relationship between folate intake and blood folate concentration was no longer linear among lactating women in our study who for the most part had been chronically exposed to high levels of synthetic folate (Figure 1). Second, the quality of the food composition tables for folate is poor and, in fact their reflection of the "actual" folate concentration in foods may have deteriorated given the variability in mandated versus the actual levels of folic acid fortification. The actual folate content of the food supply in Canada is unknown. In 1998 when Health Canada mandated folic acid fortification of the food supply, they incorporated into the legislation an allowance for overages. In the U.S. where the food supply is also fortified with folic acid, overages are estimated to range from one to two times mandated levels [36–38]. In addition, the endogenous folate content of foods as listed in the version of the CNF used in our dietary analyses was determined prior to the use of the contemporary trienzyme folate extraction procedure; hence the quality of the folate values as listed in this database remains questionable and are very likely an underestimate of the true content of many foods. However, Han et al. too recently reported weak correlations between dietary folate intake and serum (r = 0.27) and RBC folate (r = 0.29) concentrations among healthy South Korean women despite the fact that they directly measured the folate content of most folate-containing foods after trienzyme extraction [39].

Regardless of the poor predictive ability of blood folate measures in assessing relative dietary folate intake in our study, it is clear that RBC folate concentrations in this group of affluent, well-educated lactating women are generally consistent with a reduction in risk of NTD (> 906 nmol/L) as proposed by Daly and colleagues [16]. Using data from a large case-control study of 56,049 women, Daly et al reported a greater than eightfold difference in NTD risk between women with RBC folate concentrations less than 340 nmol/L compared with those with levels of 906 nmol/L or higher (P < 0.001) [16]. In fact, only two women in the present study had RBC folate concentrations = 906 nmol/L at 16-wk postpartum; interestingly both were not consuming a folate-containing vitamin supplement during lactation. The RBC folate concentrations of these two women were 869 and 822 nmol/L, which is well above cut-off values for classic folate deficiency (363 nmol/L) [20, 34].

As reported elsewhere, however, approximately one-third of this sample of lactating women had dietary folate intakes (folate endogenous to food + folic acid as a fortificant) below their estimated average requirement [40]. Assuming that folic acid is being added to the food supply at mandated levels, the synthetic folate intakes of women in the present study increased by 147 and 118 μg/d folate at 4- and 16-wk postpartum, respectively, as a result of folic acid fortification of the food supply. Together with the blood folate values reported herein, these observations suggest that as lactation continues beyond 16-wks, a greater proportion of women not consuming a folate supplement may develop RBC folate concentrations = 906 nmol/L. Subsequent to initiation of this study, Health Canada[41]., like the American Academy of Pediatrics [42] extended their recommended length of exclusive breastfeeding from 4–6 to 6 months (26 weeks). Thereafter, they recommend that infants receive complementary foods with continued breastfeeding through the first 12 months of life. In the present study, no women consuming a 400 μg/d folate supplement had a blood folate value ≤ 906 nmol/L or a folate intake below her estimated average requirement [40]. As illustrated in Figure 1, in this sample of women, there didn't appear to be an advantage of synthetic folate intakes > 151–410 μg/d in terms of maximizing RBC folate concentration. Considered together, then, these data suggest among well-nourished lactating women a folate supplement of 400 μg consumed consistently on a daily basis maximizes RBC folate content, and strikes the right balance versus higher levels (eg. 1000 or 5000 μg/d) of meeting the estimate average folate requirement of lactating women and not providing overly excessive amounts of folate than can not be incorporated into RBC precursors.

While these data are among the first to specifically assess how predictive blood folate concentrations and folate intakes are of each other post-folic acid fortification of the food supply, in addition to the quality of food composition tables, there are other limitations in estimating folate intake that must be acknowledged. Specifically, many women, particularly those that are overweight or concerned about their body weight, may under-report their dietary intake [43–45]. In addition, weighed food records, while generally regarded as the gold standard for determining dietary intakes may result in under-reporting of nutrient intake due to subject burden. Under-reporting and variability in measuring dietary intakes could have significantly weakened the statistical associations between folate intakes and blood folate concentrations in the present study. Both of the aforementioned issues could be of particular concern postpartum when many women have heightened awareness of their body weight and are busy with a new baby.

Finally, it should be noted that most lactating women in this study, on the advice of their primary care physician, consumed a 1000 μg/d supplement of folic acid during pregnancy. Many used the same folate-containing supplement prior to conception for NTD-prevention. While the life span of a RBC is ~120 days and hence by 16-wk lactation should reflect dietary and supplemental folate intakes during lactation, the contribution of liver folate stores amassed prior to parturition in meeting the nutritional requirements beyond 16 weeks of lactation could be significant. The absence of an improvement in the association between blood folate concentrations and folate intake between 4 and 16 weeks lactation could reflect this phenomenon. While the incremental increase in liver folate stores secondary to ingestion of 1000 μg/d of supplemental folic acid during pregnancy is not known, visceral folate stores in the form of long-lived pteroylpolyglutamate pools can be converted to pteroylmonoglutamates which can then be released into circulation [46, 47].

In this sample of affluent well-nourished lactating Canadian women chronically exposed to high levels of folate, plasma and RBC folate concentrations were poor predictors of relative folate intake, estimated by food composition tables generally, and synthetic folic acid intake specifically. Synthetic intakes > 151–410 μg/d in these women produced little additional benefit in terms of maximizing RBC content. More studies are needed to examine the relationship between blood folate concentration and NTD risk. Until data from such studies are available, women planning a pregnancy should continue to consume a daily folic acid supplement of 400 μg.