Response to the Draft SACN report on Folate and disease prevention

Submitted by
Anthony J A Wright and Paul M Finglas
Institute of Food Research
January 2006

SACN draft report can be found at: www.sacn.gov.uk/pdfs/report_folate_2005.pdf

• SACN largely restricted the evidence base that they considered to prospective cohort studies and randomised controlled trials in humans. In doing so they failed to consider some pertinent scientific observations, reports and hypotheses.
• New experimental research in humans, re-assessment of old research, and recent observations all combine to suggest that the anticipated exposure of the systemic blood plasma circulation to unmetabolised folic acid may have been underestimated.
• The strong impression given by SACN that B₁₂ deficiency is only correctable with high dose treatment, and the summary conclusion that “the fortification of flour with vitamin B₁₂ to improve the status in people aged 65 years and over may not be a feasible option”, is challenged.
• We argue that a wider consideration of the potential effects of mandatory folic acid fortification on the 10-30% of elderly with B₁₂ depletion/deficiency be given, rather than a narrow focus on whether the haematological clinical signs of B₁₂ deficiency due to pernicious anaemia (the minor cause of deficiency) can be ‘masked’. In particular since, the presentation of haematological abnormalities is often absent, and indeed inversely related to neurological complications, which may proceed in their absence.
• We argue that the suggested incidence of B₁₂ deficiency, and estimates of the number over 65 years of age likely to be exposed to folic acid in excess of 1,000µg/day, may have been markedly underestimated.
• We feel that plausible metabolic arguments that excess unmetabolised folic acid might either precipitate or exacerbate hypo-methylation, thus affecting inter alia the efficiency of neurotransmitter synthesis (cognition) and DNA methylation (gene expression), have not been adequately addressed.
• We feel that plausible metabolic arguments that excess unmetabolised folic acid might up-regulate dihydrofolate reductase enzyme activity, which may be accompanied by increased pyrimidine production (the rate limiting step for DNA synthesis), have also not been adequately addressed. This may potentially increase cells’ capacity for division, thus predisposing to an ‘accelerating’ effect which may be detrimental in the context of cancer.
• We draw attention to the effects of mandatory fortification on anti-folate chemotherapy, and new research showing that unmetabolised folic acid can be passed to the developing foetus – thus raising concerns of unintended influences during human embryonic development.
• In the absence of a sufficiently comprehensive picture we conclude that a true assessment of benefit / risk cannot presently be made, and that the current case for a policy of mandatory folic acid fortification encompassing the whole population is therefore unproven.
  

1. The SACN Draft Report covers a great deal of new evidence that has emerged since the last 2000 COMA Report (Department of Health, 2000) and its authors should therefore be applauded for having produced a document that is both detailed and up-to-date. However, we note that the evidence base for the Draft Report has been “largely restricted to prospective cohort studies and randomised controlled trials (RCTs) in humans.” Whilst SACN will no doubt have their reasons for this limitation (their Framework for the Evaluation of Evidence [SACN, 2002] states that “If certain types of evidence are not to be considered [eg animal or cell studies], this should be noted, giving reasons for non-inclusion”), it nevertheless appears surprising in view of the fact that much of our understanding of the absorption and initial metabolism of folate has been obtained through the use of a small-animal (rat) model. On account of the restriction in the evidence base, we believe that some pertinent scientific evidence has been overlooked. We draw attention to some specific instances of this in our Response.
   
2. Whilst it is a matter for SACN to give a final opinion/recommendation on the question of mandatory folic acid fortification, we believe that an initial “scoping consultation”, aimed mainly at researchers with an interest in folate, would have been valuable in ensuring at the outset that all significant evidence and issues potentially relating to such a policy decision were anticipated and identified.

3. SACN paragraph 13 states that “Folates are metabolised on absorption (in the gut mucosa and liver) to 5-methyltetrahydrofolate (5-MTHF), which is usually the only form found in the plasma.” Strictly speaking, this should read “Folates are metabolised on absorption (in the gut mucosa and liver) initially to 5-methyltetrahydrofolic acid, which is usually the only form found in the plasma.” [Folates are deconjugated (at the mucosal brush border) to their monoglutamate form prior to being absorbed and initially biotransformed to 5-methyltetrahydrofolic acid (the name explicitly indicating the monoglutamate structure). “5-methyltetrahydrofolate” is a generic name pertaining to an undisclosed glutamate chainlength.]
   
4. SACN paragraph 13 also states that “Oral folic acid, in excess of about 260µg, can lead to the appearance of unmetabolised folic acid in the systemic circulation (Kelly et al., 1997).” The implied corollary of this statement is that, provided that any consumption of folic acid is below this threshold, there should be no concern that unmetabolised folic acid will enter the systemic circulation to be taken up by cells and to exert any metabolic effects. This is exemplified in such statements as that promulgated by Hoffbrand (2005) in a recent ‘Symposium on Folic Acid and Health’ viz, that because “single oral doses of folic acid of less than 300µg are … converted by the gut during absorption … to … 5-methyltetrahydrofolic acid … … … fortification of the diet with levels of folic acid giving less than ca. 300µg would be predicted to have no effect on the anaemia or red cell macrocytosis due to B₁₂ deficiency.”
   
   Whilst pre-1983 there was already a reasonable degree of agreement that a significant portion of physiological doses of folic acid would undergo conversion to 5-methyltetrahydrofolic acid in the intestine (Rosenberg, 1976), it was being accepted post-1983 (following publication of the paper of Tani & Iwai, 1983) as axiomatic. However, the initial site of folic acid biotransformation in humans has been challenged recently [see following section B], leading to a re-evaluation of the Kelly et al.1997 paper in which results were obtained from acute studies in fasted volunteers that had not been subject to folic acid supplementation/fortification, the implied oral threshold (260-300µg) for the plasma appearance of unmetabolised folic acid, in the context of mandatory day-in day-out fortification, being hypothesised (eventually) to be significantly lower.

5. The watershed article of Tani & Iwai (1983) [see section above] is reflected in a general consensus that, when the mucosal extracellular folic acid concentration is low (physiological), the small intestine then absorbs and efficiently reduces and methylates folic acid and, as is the case with absorbed naturally-occurring, reduced and 1-carbon-substituted folates, subsequently transfers essentially only 5-methyltetrahydrofolic acid to the hepatic portal vein (Selhub et al., 1983; Steinberg, 1984; Mason 1990; Gregory & Quinlivan, 2002). However, the implications of differences between rat (the historical experimental animal model used in formulating most of the current understanding of folate absorption and metabolism) and man have perhaps not been fully appreciated, and the underlying scientific basis for the post-1983 consensus is now being challenged (Wright et al, 2005; Wright & Finglas 2005) on three grounds: (i) the apparent absence of a 5-methyltetrahydrofolic acid response in blood sampled directly from the human hepatic portal vein following the mucosal absorption of folic acid (Whitehead & Cooper,1967; Melikian et al., 1971), (ii) the evidence of an extremely low dihydrofolate reductase activity that seems peculiar to man (Kamen et al, 1985; Whitehead et al, 1987; Bailey et al, 2002; Bailey et al, 2003) and (iii) the implications derived from mathematical modelling of plasma labelled 5-methyltetrahydrofolic acid responses in humans to oral physiological doses of stable-isotope-labelled folates (Wright et al., 2005).
   
   If the primary initial site of folic acid metabolism in humans is actually the liver (rather than the intestinal mucosa of the upper small intestine) then, because of human liver’s known poor dihydrofolate reductase activity (needed in order to convert folic acid through dihydrofolic acid to tetrahydrofolic acid – a form that can then enter the main folate metabolic cycle), it would seem logical to predict that regular daily intake of folic acid (e.g. as a result of mandatory fortification) may eventually result in its chronic appearance in plasma of the systemic circulatory blood system. This may even happen at quite modest doses, since poor human liver dihydrofolate reductase activity may be inadequate to prevent the eventual emergence of saturation with regular intake of doses that, individually, are substantially smaller than the acute threshold dose (260 - 300 µg) reported by Kelly et al (1997) to result in the appearance of folic acid in plasma. At this point, therefore, the liver would start to fail to remove newly absorbed folic acid from the hepatic portal vein on the ‘1st-pass’, thus allowing unmetabolised folic acid into the systemic circulation. A recent report indicates that serum samples from an American population (subject to a mandatory folic acid fortification policy; 140µg folic acid /100g wheat flour) may contain a significant proportion (ca. 15% in those with a total folate concentration >22ng/ml) of unmetabolised folic acid (Pfeiffer et al., 2004). (We note in this context the calls made for the US folic acid fortification level to be raised even higher, to 350µg folic acid /100g (March of Dimes, 2005), or to be at least doubled (Brent & Oakley, 2005)).
   
6. Since SACN paragraph 6 notes that the (recalculated) UK COMA recommendation was for 280µg folic acid /100g flour (i.e. double the US level of fortification) it seems pertinent to ask what level of unmetabolised folic acid this may lead to in the systemic circulation of a UK population.
   
7. We note that SACN paragraph 7 states that “In addition to recommending folic acid fortification of flour, COMA recommended that all women should increase their folate prior to conception.” We presume that this is a call for additional self-supplementation with 400µg/day folic acid. Bearing in mind the issues raised in the paragraph above, what even greater proportion of unmetabolised folic acid may this lead to in the systemic circulation of UK females prior to conception? Further, can this unmetabolised folic acid be passed to the developing foetus? [see Response paragraph 20].

8. SACN paragraph 60 not only recognises that vitamin B₁₂ deficiency due to malabsorption is common in older people but that the majority of this is caused by the inability to release food-bound B₁₂ (associated with gastric atrophy/hypochlorhydria) rather than due to the ‘intrinsic factor’ deficiency (required for absorption) of pernicious anaemia (associated with destruction of the gastric mucosa).
   
   A complete loss of ‘intrinsic factor’ would result in a total inability to actively absorb vitamin B₁₂ no matter whether it was completely released from food-bound sources or from enterohepatically recirculated ‘free-B₁₂’. Only intramuscular injections of ‘free-B₁₂’ or passive absorption (ca. 1% of orally administered ‘high-dose’ e.g. 1,000µg/d ‘free-B₁₂’) can treat this condition. In contrast, when ‘intrinsic factor’ is still present, B₁₂ deficiency that is due only to the inability to release food-bound B₁₂ is arguably correctable with oral low-dose ‘free-B₁₂’. Indeed, because they estimated that 10-30% of older people may not be able to absorb naturally occurring vitamin B₁₂, the Institute of Medicine (2000) recommended that those over 50 years of age should meet their RDA (2.4µg) by obtaining their B₁₂ from dietary supplements (of ‘free B₁₂’).
   
   In a folic acid-fortified population, vitamin B₁₂ deficiency is the dominant nutritional cause of hyperhomocysteinemia (Green & Miller, 2005) and there have been calls for concurrent fortification with crystalline B₁₂ (Quinlivan et al., 2002; Brent & Oakley, 2005). To allow for some loss due to the bacterial overgrowth often associated with gastric atrophy/hypochlorhydria, perhaps needs could be met with low-level crystalline B₁₂ fortification that would ensure an intake of 2-3 times RDA/RNI. Such low-level fortification is unlikely to “turn flour pink” – which is the usual retort of those who still think in terms of the high-level B₁₂ fortification that would certainly be needed to address the problems of those with pernicious anaemia (the minority cause of B₁₂ deficiency).
   
9. In the light of the observations above [Response paragraph 8], SACN paragraph 66 and 67 cause concern as they give the strong impression that B₁₂ deficiency is only correctable with high-dose treatments. We cannot subscribe to the generalised SACN summary paragraph 71, which states: “A dietary intake 400 times greater than the RNI for vitamin B₁₂ would be needed to correct vitamin B₁₂ deficiency …… For this reason, the fortification of flour with vitamin B₁₂ to improve the status in people aged 65 years and over may not be a feasible option.”
   
   In line with the observations in Response paragraph 8, above, we would accept completely that SACN paragraphs 66, 67 and summary paragraph 71 would be applicable to the B₁₂ deficiency of pernicious anaemia (the minority cause of deficiency). However, given that the majority cause of deficiency is the inability to release food-bound B₁₂ (associated with gastric atrophy/hypochlorhydria), where intrinsic factor (required for absorption) remains unimpaired, and given that the Institute of Medicine (2000) infers that those over 50 years of age can meet their RDA (2.4µg) by obtaining their B₁₂ from dietary supplements (of ‘free B₁₂’), it would seem arguable that the fortification of flour with vitamin B₁₂ to improve the status in people aged 65 years and over is a feasible option.
   
10. In respect of reducing the incidence of neural tube defects (NTDs), it is of interest to note that supplementation with vitamin B₁₂ may be warranted in its own right, aside from supplementation with folate (Afman et al., 2001). These authors deduced that some mothers with NTD-affected offspring probably had a reduced affinity of the systemic plasma carrier-protein transcobalamin II, which may be explained by genetic variation in the TCII gene. Additionally, folic acid supplementation by itself may not be sufficient to minimise uracil misincorporation into DNA if serum B₁₂ concentration is low (Kalemba et al., 2005).
    
11. We note recent reviews of the link between homocysteine and cognitive function in the elderly (Morris, 2003; Garcia & Zanibbi, 2004), and the fact that the effects of cobalamin deficiency at the haematological and neurological level are not parallel, the concurrent presentation of these abnormalities being found in only 41% of cases (Healton, 1991). The Institute of Medicine (2000) noted that the major cause of clinically observable B₁₂ deficiency is pernicious anaemia, and that (unexpectedly) the occurrence of neurological complications is inversely related with the degree of anaemia, patients who are less anaemic showing more prominent neurological complications and vice versa (Healton, 1991; Savage et al., 1994), thus reinforcing the previous observation that neurological decline caused by B₁₂ deficiency can proceed in the absence of anaemia or macrocytosis (Lindenbaum et al., 1988). Although the absence of anaemia as a presenting manifestation of B₁₂ deficiency has been appreciated for some time, the range of neurological and psychiatric manifestations continues to expand (Alpers, 2005). Some patients can manifest neurological symptoms of B₁₂ deficiency when serum total B₁₂ is in the normal range, even up to 350pg/ml (Alpers, 2005).
    
    Since the majority of B₁₂ deficiency that is due only to the inability to release food-bound B₁₂ may be insidious, taking many years to progress eventually (if at all) to pernicious anaemia, one wonders how much cognitive decline (or deterioration in the efficiency of other methylation-dependant processes connected with 1-carbon metabolism e.g. DNA-methylation associated with gene expression) might conceivably develop in the absence of clinically-observable signs of B₁₂ deficiency. In considering mandatory folic acid fortification, is it therefore appropriate to focus only upon the risks of ‘masking’ the haematological clinical signs of B₁₂ deficiency which may (sometimes, but not always) be associated with pernicious anaemia – rather than asking what may be a wider and more pertinent question: “what are the potential effects of mandatory folic acid fortification on the 10-30% of elderly with B₁₂ depletion/deficiency, particularly where the main cause will be the inability to release food-bound B₁₂, rather than loss of intrinsic factor”? Might an appreciable concentration of systemically-circulating unmetabolised folic acid precipitate or exacerbate aberrant neurotransmitter synthesis, potentially causing neurological damage without inducing clinical signs? Are there also unevaluated risks of precipitating or exacerbating aberrant DNA-methylation and gene expression?
    
12. We note the theory of cerebral oxidative-stress-induced hyperhomocysteinemia and the suggestion that currently available pharmaceutic forms of vitamin B₁₂ are unlikely to be efficiently utilised by neurons under these conditions, and that the use of glutathionylcobalamin (or concurrent B₁₂ in combination with a glutathione precursor such as N-acetylcysteine, NAC) might be preferential as a treatment (McCaddon et al, 2002; McCaddon & Davies, 2005). A clinical trial of high-dosages of B₁₂ + NAC + folic acid is currently being conducted in the US to formally evaluate the clinical responses seen in the case studies [Andrew McCaddon – personal communication].
    
13. SACN paragraph 62 acknowledges that plasma holotrancobalamin (holoTCII) concentrations, rather than total serum B12 ± plasma total homocysteine (tHcy) ± plasma methylmalonic acid (MMA) concentrations, may be a sensitive diagnostic indicator of vitamin B₁₂ deficiency. HoloTCII has been shown to have no significant circadian variation (Hvas et al., 2005) and shows promise as a first-line test for diagnosing early vitamin B₁₂ deficiency (Hvas & Nexo, 2005). Only that proportion (5-20%) of total serum B₁₂ that circulates as holoTCII can be delivered and incorporated into cells of the body. On the basis of albeit limited evidence, a low holoTCII concentration (which may be the earliest indicator of negative cobalamin balance and warning of biochemical deficiency) may be extensive in the elderly. Flynn et al. (1997) reported that whilst only 7% of 171 apparently healthy individuals (mean age 65 years; range 41 to 85) exhibited a suboptimal serum total B₁₂ concentration, 49% exhibited little if any absorption of dietary B₁₂ – as indicated by suboptimal holoTCII concentration.
    
14. SACN paragraph 64 refers to the work of Clarke et al (2004) who assessed B₁₂ deficiency as between 5 and10% based either on a B₁₂<150pmol/l, or serum B₁₂<200pmol/l and total homocysteine >20µmol/l. It was these figures that were used later [SACN paragraph 111] to calculate the number of UK people having B₁₂ deficiency who would be exposed to folic acid in excess of 1000µg/day if the mandatory fortification of flour with 240µg/100g were introduced (Department of Health, 2000). However, diagnostic algorithms using total serum B₁₂, homocysteine and methylmalonic acid measurements have been dismissed recently as totally unreliable for the early recognition of B₁₂ deficiency, where it was concluded that they neither predict nor preclude the presence of cobalamin-responsive haematologic or neurologic disorders (Solomon, 2005). Indeed, of the 456 patients who were evaluated for B₁₂ deficiency, it was determined that if therapy had been restricted to symptomatic patients with both low or intermediate B₁₂ levels and increased metabolite values, 63% of responders to subsequent B₁₂ treatment would not have been treated. This suggests that the assessment of B₁₂ deficiency in the UK population (Clarke et al., 2004) may have been underestimated by a factor of two or more.

15. SACN paragraph 111 calculates the number of UK people aged 65+ having B₁₂ deficiency that would be exposed to folic acid in excess of 1000µg/day if the mandatory fortification of flour with 240µg/100g were introduced (Department of Health, 2000; Table A7.2, page 100; i.e. 0.8% of men and 0.5% of women) based upon the figures of Clarke et al (2004) who assessed UK B₁₂ deficiency as between 5 and 10% for those aged 65+ years. The figures put forward were:- “an estimated 2921 (1564 men, 1357 women) to 5844 (3129 men, 2715 women) of people would be at risk…..” However, bearing in mind that it has been estimated that the US population was actually exposed to double the projected intake of folic acid, we do not think it unreasonable to ask what the UK projected exposure figures would be if the (corrected) COMA recommendation for folic acid at 280µg/100g were increased by only 50% (to the equivalent of 420µg/100g). The tabulated estimates for men then rise to 10.4% and for women to 2.2%. On this basis, an estimated 26300 (20337 men, 5973 women) to 52619 (40674 men, 11945 women) of people would then be at risk. As argued in Response paragraph 14, the assessment of B₁₂ deficiency in the UK population (Clarke et al., 2004) may have been underestimated by a factor of two or more. Therefore, even these revised estimates of the exposure of B₁₂-deficient people to folic acid in excess of 1000µg/day may be underestimated by a factor of two. Furthermore, B₁₂ deficiency does not just start at age 65. Additionally, no allowance has been made for future demographic changes and the prediction that the number of people aged 65 and over is predicted to increase by about 53% between 2001 and 2031 (Majeed & Aylin, 2005).

16. Efficiency of neurotransmitter synthesis and DNA methylation: There is a plausible metabolic argument (Scott & Weir, 1998) that excess folic acid may either precipitate or exacerbate hypo-methylation. A deficiency of B₁₂ (required for de novo re-synthesis of methionine from homocysteine) might compromise both neurotransmitter synthesis (which requires donation of a methyl group from methionine; regenerating homocysteine in the process) and DNA methylation (which regulates gene expression). It might also retard DNA synthesis and repair because folate will become ‘trapped’ as 5-methyltetrahydrofolate and this will diminish the availability of the 1-carbon-substituted forms of folate (10-formyltetrahydrofolate and 5-10- methylenetetrahydrofolate) that are required for the synthesis of purines and pyrimidines. This would be expected, in turn, to reduce the rate of cellular turnover. Unmetabolised folic acid can relieve any metabolic reduction or block in the DNA synthesis required for cell turnover, but not any reduction in the de novo re-synthesis of methionine from homocysteine caused by B₁₂ deficiency. However, new cells require both structural and functional proteins as well as DNA. Competition for the (reduced) methionine pool between the demands of methylation reactions and protein synthesis may be an important consideration since this might precipitate or exacerbate hypomethylation, leading possibly to detrimental impacts on gene expression, neurotransmitter synthesis and cognitive ability. Unexpectedly, in an American population subject to a mandatory folic acid fortification policy, a high intake of folate has been associated with accelerated cognitive decline in older persons, leading to the suggestion that this may be due to an exacerbation of the effects of functional B₁₂ deficiency (Morris et al., 2005), which has been estimated to be 10-30% in the elderly (Institute of Medicine, 2000). A review, with historical perspective, of the benefits and risks of folic acid to the nervous system has been published (Reynolds, 2002).
    
17. Effect on Cancer: It is important here to distinguish between the relationship between (naturally-occurring) folate and cancer and the potential effects of appeciable concentrations of systemically-circulating unmetabolised folic acid (pteroylmonoglutamic acid) and cancer. Low folate status may or may not be a risk factor for (some) cancers. However, as early as 1947, it was observed that treatment of children with acute leukaemia with folic acid could result in an “acceleration phenomenon” of the cancer (Farber at al., 1948; Farber, 1949). In these cases, folic acid was injected either intramuscularly or intravenously, and the dose was supra-physiological (Farber et al., 1947). However, we must not lose sight of the fact that it was explicitly the observations that folic acid therapy enhanced the growth of tumours that led directly to research into folate antagonists (Hoffbrand & Weir, 2001).
    
    As compared to the activity in fresh human cells ex vivo, similar cells cultured in vitro with folic acid as the source of folate can have their dihydrofolic acid reductase (DHFR) activity increased by 100-fold or more (Kamen et al., 1985). An up-regulation of DHFR activity may be accompanied by an increased thymidylate synthase activity since the transcription of both these genes is co-regulated by the same transcription factor. Mathematical modelling indicates that this would increase pyrimidine production (the rate-limiting step for DNA synthesis) without significantly affecting the rest of folate metabolism (Nijhout et al., 2004). It could thus be hypothesised that, in contrast to an increased exposure to the naturally circulating folate [6S]-5-methyltetrahydrofolic acid, exposure to unmetabolised folic acid may increase cells’ capacity for division, thus predisposing to an ‘accelerating’ effect which is detrimental in the context of cancer.
    
    Detrimental effects on cancer risk were reported from the NORVIT randomised trial of 3,749 Norwegian myocardial infarction patients, presented at the 2005 Congress of the European Society of Cardiology, based on post-hoc analyses of those subjects (split 50:50) who did or did not receive folic acid (Bonna, 2005). It caused concern that results suggested a statistical trend for an overall 40% increase in the relative risk of previously undisclosed cancers being identified in the groups taking folic acid when compared to control groups [RR 1.4 (95% CI, 1.0 – 2.0), P=0.08]. Animal studies have suggested that folic acid supplementation may accelerate tumour progression if too much is given or if it is provided after neoplastic foci are established (Kim, 2004). A greater number of colonic aberrant crypt foci have been observed in rats supplemented with folic acid (Buckley et al., 2005).
    
    Currently, there would probably be a consensus that there was a paucity of direct evidence that appeciable concentrations of systemically-circulating unmetabolised folic acid may be a risk factor for cancer ‘acceleration’. However, the danger here is that this may be because this question has not been addressed rigorously. In this context, the mantra ‘absence of proof is not proof of absence’ may justifiably be considered appropriate.
    
18. Effect on anti-folate chemotherapy: A further area of concern is the potential for negative effects on chemotherapy using the anti-folate drug, methotrexate, whose mode of action is to limit folate availability to cells by reducing (by substrate competition) the activity of dihydrofolate reductase – an enzyme whose gene could also conceivably be up-regulated through exposure to high concentrations of folic acid. Post-hoc analysis from two randomised, controlled studies has indicated that folic acid reduces the degree of improvement of methotrexate-treated rheumatoid arthritis patients (Khanna et al., 2005). Additionally, a recent study in a US population has indicated that high serum folate levels above ~50nmol/L (~22ng/ml; a similar concentration above which unmetabolised folic acid makes up ca. 15% of total folate: see Response paragraph 5) significantly increases the failure of ectopic pregnancy treatment with single-dose methotrexate (Tacaks & Rodriguez, 2005).
    
19. Effect on folate metabolism: Folic acid is not a natural coenzyme and the long-term biological effects of exposure are unknown. Whilst accepting that such exposure may present no health risk at all Lucock (2005) has commented that we cannot know this for certain and raised a number of research issues that should be addressed, including an assessment of the in vivo effect of folic acid on all folate-dependent enzymes.
    
20. Can unmetabolised folic acid pass to the developing foetus?: Maternal concentrations of folate predict values in foetal blood at delivery (Obeid et al., 2005), but the question of interest is whether unmetabolised folic acid from fortified foods and supplements can be passed to the developing foetus. In mothers not subject to mandatory folic acid fortification, none of whom were consuming folic acid supplements in the latter stages of pregnancy but who were potentially exposed to dietary intakes of folic acid from commercially available fortified foods, there has been recent evidence of unmetabolised folic acid in the cord blood of newborn, and an increase of unmetabolised folic acid in serum of 4-day-old infants, post formula-feeding (Sweeney et al., 2005).
    
21. Can folic acid modify gene methylation and gene expression in a differential (and possibly deleterious) manner to 5-methyltetrahydrofolic acid – the natural circulating folate form (a) in the developing foetus, and (b) in general?: From results using a mouse model it was concluded that population-based supplementation with folic acid, intended to reduce the incidence of neural tube defects, may have unintended influences on the establishment of epigenetic gene-regulatory mechanisms during human embryonic development (Waterland & Jirtle, 2003).
    
    Surprisingly, we cannot find any research that has looked at the effects of folic acid supplementation on global gene expression in humans, and yet this would be so easy to do in the UK population where one could supplement volunteers at 400µg/day (and additionally monitor the concentration of systemically circulating unmetabolised folic acid with increasing time).
    
    Again in a mouse model, dietary folic acid supplementation has been reported to accelerate CpG island methylation of the putative tumor suppressor gene ESR1, gene-silencing being increasingly recognised as an important phenomenon in colorectal carcinogenesis (Belshaw et al., 2005).
    
22. Effect on genetic selection: A concern has been raised that exposure to supplemental folic acid could induce a strong genetic pressure – one that has the side effect of increasing the prevalence of a ‘common’ methylenetetrahydrofolate reductase (MTHFR) C677T gene variant resulting in the birth of more ‘TT’ newborns who may be more susceptible to chronic diseases if then subsequently exposed to low dietary folate intake (Lucock & Yates, 2005).

23. We accept the evidence that a mandatory policy of folic acid fortification will reduce the risk of NTDs. However, the potential detrimental effects of the systemic circulation of unmetabolised folic acid (which may be much higher than previously assumed) have not been seriously addressed, possibly because it has always been assumed that there would not be any! There is such a paucity of good quality research at the fundamental level, particularly addressing the putative effects of the systemic circulation of unmetabolised folic acid on gene expression, post-transcriptional regulation and metabolism, and the interrelationships with B₁₂ status, that we do not believe a soundly-based benefit / risk analysis could currently be undertaken.
    
    A rigorous benefit / risk analysis of mandatory folic acid fortification must include age-subgroups (foetus, babies, infants, children, adolescents, adults, elderly and the aged) and both genders of the population. Also, it cannot afford to omit potentially important theoretical or observational concerns just because they have not been scientifically fully addressed: A ‘weighting’ may be able to be given to such factors, based on a judgement of the relative risks involved. Whilst recognising the potential benefits of mandatory folic acid fortification, we should also recognise, and address, any potential limitations – even if currently they are mainly theoretical. The effect of folic acid on DNA methylation at epigenetically susceptible loci in the embryo and foetus needs to be addressed in order to rule out unintended influences on epigenetic gene-regulatory mechanisms during human embryonic development.
    
    In an age of demographic change resulting in an increasingly aged population with maybe high levels of functional vitamin B₁₂ deficiency, the potential for folic acid fortification to exacerbate neurological symptoms such as cognitive decline must be taken seriously as, of course, must the rate of functional B₁₂ deficiency itself. In this respect priority needs to be given to accurate assessments of the prevalence of functional B₁₂ deficiency.
    
    We strongly counsel that all potentially important unintended effects that could result from the systemic circulation of unmetabolised folic acid should be addressed methodically to ascertain a true picture of benefit / risk of fortification (and self-supplementation) with folic acid. This could/should be contrasted to an assessment of the benefit / risk for an equivalent fortification (or self-supplementation) with [6S-]-5-methyltetrahydrofolic acid, the natural circulating folate form.
    
    In the absence of a sufficiently comprehensive picture we conclude that a true assessment of benefit / risk cannot presently be made, and that the current case for a policy of mandatory folic acid fortification encompassing the whole population is therefore unproven.

24. Animal studies: The Framework for the Evaluation of Evidence (SACN, 2002) [Part B – Assessment: Animal studies part (f)] notes that an assessment should be undertaken of the “Comparability of dietary exposures to human dietary levels (in UK/Europe)”. However, we do not accept this to be directly applicable in the context of response to folic acid. This is because animals, unlike humans (who appear to be unique) [See Response paragraph 5], have a comparatively high dihydrofolic acid reductase (DHFR) activity. Consequently, if assessing the impact of systemic exposure, animals would de facto have to be orally dosed with a much greater than pro-rata amount of folic acid in order to elicit the same circulating serum/plasma concentration of unmetabolised folic acid.
    
25. Cell culture studies: The Framework for the Evaluation of Evidence (SACN, 2002) [Part B – Assessment: Cellular and molecular studies part (d)] notes that an assessment should be undertaken of the “Extent to which data from cell studies is likely to be relevant”. Researchers need to exercise extreme care when using cell culture studies to study the putative effects of folic acid – particularly on gene expression – because it has been noted (Kamen et al., 1985) that dihydrofolate reductase (DHFR) has an activity in normal human tissue, and human tumour and leukaemia cells in vivo, that is several magnitudes lower than present in human cell lines grown in vitro. The hypothesis put forward for this discrepancy is that the higher level of DHFR activity found in in vitro cell lines could be due to the high levels of folic acid used historically as the source of folate in culture media. Therefore, we caution that in vitro cell lines need to have been re-grown/acclimatised using [6S-]-5-methyltetrahydrofolic acid, the natural circulating folate form, until analysis has indicated that DHFR activity (in particular) has returned to a normal level. Only then can any effect of folic acid on gene expression (particularly DHFR) be accepted provisionally as valid.