22a.1 Folate

Folate is a generic term referring to both natural folates and folic acid (pteroyl­monoglutamic acid, PGA), the synthetic form used in supplements and fortified food. All folate forms comprise three moieties: a pteridine; a p‑amino­benzoic acid (PABA) and a glutamate residue (Figure 22a.1).

The parent  compound, PGA, is completely oxidized and not found in nature. The natural folate forms are reduced molecules, with the addition of 2 or 4 hydrogen atoms to the pteridine, giving rise to dihydro­folate or the various tetra­hydro­folate (THF) forms. THF can carry one-carbon groups attached at the N‑5 (methyl, formyl or formimino), the N‑10 (formyl) or bridging N‑5 and N‑10 (methylene or methenyl) positions of the pteridine ring, giving rise to a number of different cofactor forms of folate. Also, whereas folic acid is a mono­glutamate, containing only one glutamic acid residue, most natural food folates exist as poly­glutamate derivatives containing additional glutamate residues bound in peptide linkage to the gamma-carboxyl group.

22a.1.1 Functions of folate

Folate coenzymes are required for one-carbon meta­bolism involving the transfer and ulilization of single-carbon atom units, including methylene (CH₂), forminino (CH=NH), methyl (CH₃), methenyl (CH), and formyl (CHO) groups (Figure 22a.2).

Folate is thus essential for key biolog­ical functions, including, serine and glycine meta­bolism, histidine cata­bolism, methionine synthesis and in thymidylate and purine bio­synthesis, precursors of DNA. Folate, in the form of 5‑methylTHF, along with vitamin B₁₂, is required for the synthesis of methionine from homo­cys­teine (catalyzed by the enzyme methionine synthase), and in turn, the synthesis of S‑adenosyl­methionine (SAM). This methyl group donor is used in numerous biolog­ical methy­lation reactions, including the methy­lation of a number of sites within DNA, RNA, proteins, and phospho­lipids. The reader is referred to Bailey et al. (2015) for a detailed review of the functions of folate.

Folate inter­action with other B vitamins. For folate to function within one-carbon meta­bolism, it interacts closely with vitamin B₁₂, vitamin B₆ and ribo­flavin (McNulty et al., 2019). Reduced folates enter the one-carbon cycle as THF which acquires a carbon unit from serine in a vitamin B₆‑dependent reaction to form 5, 10 methyleneTHF. This co-factor form, once formed either converted to 5‑methylTHF or serves as the one-carbon donor in the synthesis of nucleic acids, where it is required by thymidylate synthetase in the conversion of deoxy­uridine (dUMP) to deoxy­thymidine mono­phosphate (dTMP) for pyrimidine biosynthesis, or is converted to other folate forms required for purine biosynthesis. Methylene­tetra­hydro­folate reductase (MTHFR) is a ribo­flavin-dependent enzyme that catalyzes the reduction of 5, 10 methyleneTHF to 5 methylTHF, the folate form used by methionine synthase for the vitamin B₁₂-dependent conversion of homo­cys­teine to methionine and the formation of THF. Methionine is activated by ATP to form S-adenosyl­methionine (SAM), which then donates its methyl group to more than 100 methyl­trans­ferases for a wide range of substrates such as DNA, hormones, proteins, neuro­trans­mitters and membrane phospho­lipids, all of which are regulators of important physio­logical processes (Bailey et al., 2015). In summary, effective folate functioning requires essential meta­bolic inter­action with vitamins B₁₂, B₆ and ribo­flavin. Thus sub-optimal status of one or more of these B vitamins, or poly­morphisms in folate genes, can impair one-carbon meta­bolism, even if folate status is sufficient.

Absorption and transport of folate. Folate absorption takes place by an active process, primarily from the proximal part of the jejunum. Before absorption, the poly­glutamate forms of folate are decon­jugated to the mono­glutamate form in the gut by the intestinal brush-border enzyme glutamate carboxy­peptidase II (GCPII), more commonly termed “folate conjugase”. The mono­glutamates are then taken up by the mucosal cells. Folates require transporters to cross cell membranes; these include the reduced folate carrier (RFC), the proton-coupled folate trans­porter (PCFT), and the folate receptor proteins, FRα and FRβ. Folic acid is a mono­glutamate and thus does not require decon­jugation before uptake by intestinal cells.

Most of the folate mono­glutamates absorbed from the gut are transported to the liver (the major storage organ for folate) and are reconverted intra­cellularly within hepato­cytes to poly­glutamate derivatives by the enzyme folyl­poly­glutamate synthetase. These folate poly­glutamates are stored in the liver or converted to 5‑methyl­tetra­hydro­folate for secretion into the bile, and then reabsorbed by way of the entero­hepatic circulation. This recircu­lation process may account for as much as 50% of the total folate that reaches the periph­eral tissues. Circulating folate in blood is found in the mono­glutamate form, predominantly as 5‑methyl­tetra­hydro­folate.

22a.1.2 Deficiency of folate in humans

Severe deficiency of folate leads to megalo­blastic anemia, presenting as fatigue, weakness, and shortness of breath owing to a low red blood cell count. This condition is hemato­logically characterized by the presence of immature, enlarged nucleated cells, reflecting impaired DNA synthesis as a result of folate depletion. Rapidly prolif­erating cells are especially sensitive to abnormalities in DNA synthesis. Hence, the manifes­tations of folate defi­ciency appear first in the hemato­poietic system and then in the epithelial cell surfaces and the gonads. Abnormal cell replication in the hemato­poietic system, manifested by hyper­segmented neutro­phils, is one of the earliest morpho­logical changes. Later, megalo­blasts appear in the bone marrow and macro­cytes in the peripheral blood. Some additional signs and symptoms that have been reported with clinical folate defi­ciency include fatigue, angular cheilosis, anorexia, insomnia, glossitis, recurrent aphthous ulcers, and pallor of the skin and mucous membranes. Of note, the megalo­blastic anemia resulting from folate defi­ciency is identical to that resulting from vitamin B₁₂ defi­ciency, and therefore specific bio­marker testing is essential to provide a differential diagnosis (See Section 22a.2).

Causes of low and deficient folate status. Folate defi­ciency arises from various causes, relating either to increased requirements, reduced avail­ability, or both. Pregnancy is a time when folate requirement is greatly increased in order to sustain the demand for rapid cell replication and growth of fetal, placental and maternal tissue (McNulty et al., 2019). Gastro­intestinal conditions, such as celiac disease, can also lead to deficient folate status through chronic malab­sorption. Certain drugs, including phenytoin and primidone (anti­convulsants) and sulfa­salazine (used in inflammatory bowel disease), are also associated with folate defi­ciency through adversely affecting folate meta­bolism. Like­wise, heavy alcohol consumption and smoking are also linked with lower folate status (Bailey et al., 2015).

Prevalence of defi­ciency. Folate defi­ciency occurs commonly in pregnant and lactating women in low- and middle-income countries, where it is reported to occur in > 20% of women of repro­ductive age (Rogers et al., 2018). Here, dietary intakes of folate are often inadequate to meet the high requirements of pregnancy.

In high-income countries, clinical folate defi­ciency is less common (due largely to the beneficial effects on folate status of supple­men­tation and/or food fortifi­cation), but subclinical defi­ciency (indicated by low serum and RBC folate concen­trations) is widespread, particularly in women of reproductive age, especially if pregnant and lactating, in European countries. Low folate status has also been reported in low birth weight and premature infants (Scholl & Johnson, 2000) and adolescents of low socio­economic status (Bailey et al., 2015). Serum and RBC folate concen­trations typically decrease throughout pregnancy; however, supple­men­tation with folic acid prevents this decline and can thus prevent the occurrence of megalo­blastic anemia of pregnancy (Blot et al., 1981; McNulty et al., 2013). Estimates of folate defi­ciency can however vary consid­erably across different populations and population sub-groups, depending on the method used and the cutoff points applied to bio­marker measures of folate status (Bailey et al., 2015).

Neural tube defects. Maternal folate nutrition before and in early pregnancy is known to play a critical role in fetal development. Notably, conclusive scientific evidence published 30y ago shows that folic acid supple­men­tation with folic acid (the synthetic form of the vitamin) in early pregnancy protects against both first occurrence (Czeizel & Dudás, 1992) and recurrence (MRC, 1991) of neural tube defects (NTDs). These major birth defects occur as a result of a failure of the neural tube to close properly in the first few weeks of pregnancy, leading to death of the fetus or newborn, or to various disabilities involving the spinal cord, the most common form of which is spina bifida. The conclusive evidence that folic acid can prevent NTD has led to clear folic acid recommendations for women of reproductive age which are in place worldwide. It is important to appreciate, however, that the risk of NTD is increased when maternal folate status is low, but not necessarily within the range typically classed as folate defi­ciency. In a large prospective study of women in Ireland (where rates of NTD are among the highest in the world), a woman's risk of having a child with an NTD was found to be strongly associated with pregnancy RBC folate in a continuous dose-response relation­ship (Daly et al., 1995).

In practice, implementing folic acid recommendations so that women and their babies benefit, is somewhat challenging. Although mandatory folic acid fortif­ication of foods has proved to be highly effective in reducing NTD wherever it has been introduced (in 84 countries worldwide), elsewhere preventable NTDs are not being prevented including in European countries. Notably, one recent study estimated that from 1998 to 2017, a total of 95,213 NTD pregnancies have occurred amongst 104 million births in 28 European countries; a prevalence of 0.92 per 1,000 births (Morris et al., 2021). This study concluded that failure to implement mandatory folic acid fortif­ication in the 28 European countries continues to cause NTD to occur in almost 1,000 preg­nancies every year.

The precise mech­anism explaining the beneficial effects of peri­concep­tional folic acid against NTD remains uncertain, though proposed mech­anisms have focussed on factors that impair normal folate meta­bolism, including poly­morphisms in folate genes. Among the latter, an increased risk of NTD is most strongly associated with the 677C>T variant in the gene encoding the folate-meta­bolising enzyme methylene­tetra­hydro­folate reductase (MTHFR), as reported in most studies and meta-analyses (Botto & Yang, 2000; Vollset & Botto, 2004). Auto­anti­bodies against folate receptors are also implicated in preg­nancies affected by NTD (Rothenberg et al., 2004). Also, although low maternal folate is considered the major contributing factor in NTD, convincing evidence shows that low vitamin B₁₂ is an independent risk factor in NTD (Molloy et al., 2009). In addition, apart from preventing NTD, there is good evidence that peri­concep­tional folic acid use may prevent congenital heart defects in infants (van Beynum et al., 2010), and possibly orofacial clefts (Bailey et al., 2015).

22a.1.3 Food folate sources, bio­avail­ability and dietary intakes

Folates are widely distributed in animal and plant foods, but only certain foods are rich sources, including liver, yeast, green leafy vegetables, asparagus, beans, legumes, where they exist primarily as poly­glutamates, containing up to nine glutamate residues attached to the p‑amino­benzoic group of the molecule (Figure 22a.1). In contrast, folic acid, the synthetic vitamin, is a mono­glutamate, containing only one glutamic acid residue in its structure. Also, unlike folic acid, which is a fully oxidized molecule, natural folates are reduced at the 5, 6, 7 and 8 positions of the pteridine ring and so are prone to oxidative cleavage at the C9‑N10 bond producing two degradation products, a pteridine and p‑amino­benzoyl­glutamate, both of which are inactive and cannot be biolog­ically converted to any active folate form.

Folate bio­avail­ability. Bioavail­ability refers to the proportion of ingested nutrient that is absorbed and available for meta­bolic processes. Naturally-occurring food folates are only partially bioavail­able (McNulty & Pentieva, 2004). Natural food folates are reduced molecules, and therefore are inherently unstable outside living cells. In addition, the ease with which folates are released from different food matrices and their conversion to the mono­glutamate form before uptake by intestinal cells can vary greatly. Folate bio­avail­ability from different food sources is also dependent on the presence of certain dietary constituents, that may enhance folate stability during digestion, or inhibit bio­avail­ability owing to specific inhibitors of deconjugation. Thus, the bio­avail­ability of food folates from a mixed diet is limited and highly variable. In addition, natural food folates (particularly green vegetables) can be unstable during cooking, and this can substantially reduce the folate content of a food before it is ingested (McKillop et al., 2002). This is an additional factor that further contributes to the limited potential for natural food folates to positively influence folate status (Cuskelly et al., 1996).

Dietary folates. In contrast to the natural folate forms which have limited stability and bio­avail­ability, folic acid provides a highly stable and bio­available vitamin form. The bio­avail­ability of folic acid is assumed to be 100% when ingested as a supple­ment, while folic acid in fortified food is estimated to have about 85% the bio­avail­ability of supple­mental folic acid (Pfeiffer et al., 1997). These differences have led to the develop­ment of “dietary folate equivalents” or DFE values; more details are given in Section 8a. Briefly, expressing dietary folate intakes and recommend­ations in DFE terms, allows an adjust­ment for the differences in bio­avail­ability between natural food folates and the synthetic vitamin. The DFE is defined as the quantity of natural food folate plus 1.7 times the quantity of folic acid in the diet; this definition is based on the assumption that the bio­avail­ability of folic acid added to food is greater than that of natural food folate by a factor of 1.7 (Institute of Medicine, 1998). This estimation is largely dependent on a meta­bolic study in non-pregnant women that estimated the bio­avail­ability of food folates to be 50% relative to that of folic acid (Sauberlich et al., 1987), and other evidence, mentioned above, showing that folic acid added to food has 85% of the bio­avail­ability of free folic acid (Pfeiffer et al., 1997).

Fortified foods as a source of folate. Food fortif­ication, the process of adding essential micro­nutrients to food, plays an important role in facilitating a more optimal nutritional status in individuals and populations. In the case of folate, it plays a crucial role. Folic acid, the folate form used for food fortif­ication, is cheap to produce, very stable once added to foods and highly bioavailable when ingested. Thus, depending on national fortif­ication policy and/or access to folic acid-fortified foods, the folate status of populations can vary greatly from one country to the next This is reflected in differences in health outcomes (notably in relation to NTD risk). Food fortif­ication may be under­taken on a voluntary or mandatory basis. Voluntary fortif­ication, whereby folic acid is added to foods such as break­fast cereals at the discretion of the manufac­turer, is permitted in most European countries. This results in higher folate intake and status (Hopkins et al., 2015), but the benefit is limited only to consumers who choose to eat the fortified food products. However, when fortif­ication is undertaken on a mandatory (population-wide) basis, including in the USA, Canada, Australia and Chile, it has proved to be highly effective, not only in increasing folate status and reducing folate defi­ciency (Yang et al., 2010), but also in reducing NTD in that country (Honein et al., 2001, Lopez-Camelo et al., 2005, De Wals et al., 2007, Sayed et al., 2008). Mandatory fortif­ication is now in place in 85 countries worldwide, both high‑ and low-middle income countries

Effects of high intakes of folate. High folate intakes are generally not assoc­iated with any adverse effects. However, there are concerns of potential adverse effects of excess intakes of folate acid, the synthetic vitamin form. Excessive folic acid intake consti­tutes exposure doses that exceed the Tolerable Upper Intake Level (UL) of 1000µg/d for adults, as set by the U.S. Institute of Medicine (1998).

Historically the concern regarding excess folic acid focused on the potential to mask pernicious anemia and exacerbation of the clinical effects of vitamin B₁₂ defi­ciency. More recently other concerns have been raised, including potential adverse effects on cancer risk, birth outcomes, and other diseases. However, a recent report from a 2019 expert workshop tasked with reviewing this research area, concluded that there is an insufficient body of evidence to support adverse human health outcomes as a result of high intakes of folic acid. None­theless, these experts called for further research to determine the safety of excess folic acid intake (Maruvada et al., 2020).

22a.2 Biomarkers of folate status

In 2015, an expert international panel produced a comprehensive review of folate biology and bio­markers as part of the Biomarkers of Nutrition for Development (BOND) project convened by the U.S. National Institutes of Health (Bailey et al., 2015). This review identified serum folate, red blood cell (RBC) folate and plasma homo­cys­teine as the “Priority Folate Biomarkers” for assessing folate status. Serum folate reflects recent dietary intake. RBC folate, compared with serum folate, is a better indicator of folate intake and status over the previous 3–4mos. Plasma homo­cys­teine provides a functional indicator of folate status and is elevated in folate defi­ciency, on the basis that normal homo­cys­teine meta­bolism requires an adequate supply of folate. Concentrations of folate in erythro­cytes, but not serum, also fall in vitamin B₁₂ defi­ciency. Ideally, both serum and RBC folate concen­trations should be measured.

Cut-off values for assessment of folate status. The sequential stages in the development of folate defi­ciency were originally estab­lished through detailed depletion / repletion experiments conducted during the late 1960's, providing a basis for setting cut-off values for serum and RBC folate for the assess­ment of status (Herbert, 1987); these sequential stages of folate defi­ciency and related cut-off values are summarized in Box 22a.1 The initial stage of folate defi­ciency, termed negative folate balance, is consistent with serum folate values of < 7nmol/L (Herbert, 1987). If the negative balance persists, tissue folate becomes depleted, as indicated by RBC folate falling below the normal range to < 363nmol/L. At this second stage, there is little evidence that biochemical function is impaired, although plasma homo­cys­teine concen­trations may be slightly elevated. By the third stage, termed folate-deficient erythropoiesis, functional impairment is usually evident, with erythrocyte folate values < 272nmol/L. Tissue folate stores are severely depleted in the fourth and final stage where the classical hematological changes occur, manifested as folate-defi­ciency anemia. These include macro-ovalocytic erythro­cytes in the circulating blood and megaloblasts in the bone marrow. At this stage, hypersegmented neutro­phils in the peripheral blood smear and abnormal RBC indices are also apparent; mean red cell volume and mean cell hemo­globin are elevated and hemo­globin concen­tration is low.

More recently, revised cut-off points for folate defi­ciency (serum folate <10nmol/L and RBC folate < 340nmol/L) have been defined based on the meta­bolic indicator, plasma homo­cys­teine (Selhub et al., 2008), as summarized in Box 22a.1.

22a.2.1 Serum and RBC folate concen­trations

Serum folate represents the sum of several folate forms circulating in blood. The main circulating folate is 5‑methyl‑THF; other reduced forms such as THF and formyl‑THF may also be present, but in very small concen­trations (Bailey et al., 2015). As will be discussed later (section 22a.2.4), unmeta­bolized folic acid can also be present in varying concen­trations in plasma (Bailey et al., 2010).

Red cells contain much higher folate concen­trations than serum. 5‑methyl‑THF poly­glutamates are the main RBC folate forms. The measure­ment of RBC folate is more complex than that of serum folate, because of the need to convert poly­glutamates to mono­glutamates prior to analysis. In individuals homo­zygous for the MTHFR C677>T poly­morphism, a portion of the 5‑methyl‑THF poly­glutamates is replaced by formyl-folates (Bagley & Selhub, 1998).

Serum folate is the earliest indicator of altered folate exposure and reflects recent dietary intake (Pfeiffer et al 2010). In the individual, serum folate concen­trations can increase markedly in response to dietary folate, reaching a peak concen­tration 90min after ingestion.

RBC folate is a sensitive indicator of long-term folate status. RBC folate parallels liver concen­trations, accounting for about 50% of total body folate, and thus reflects tissue folate stores (Wu et al., 1975). RBC folate, compared with serum folate, responds more slowly to changes in dietary folate intake and is a better indicator of folate intake over the previous 3–4mos when circulating folate is incorporated into the maturating red cells during erythropoiesis, thereby reflecting folate status during the preceding 120d, the half‑life of the RBC (Mason, 2003).

Factors affecting serum and RBC folate concen­trations

Dietary folate intakes. Blood folate levels are primarily affected by dietary folate intakes. Serum and RBC folate are highly responsive to intervention with folic acid (Duffy et al., 2014), with natural food folates typically resulting in poorer folate responses compared to folic acid at similar intervention levels (Cuskelly et al., 1996). Like­wise, population data show that both serum and RBC folate concen­trations are highly reflective of exposure to folic acid, with the highest concen­trations observed in people who consume folic acid in both supplements and fortified foods (Yang et al., 2010; Hopkins et al., 2015).

Vitamin B₁₂ status. Because vitamin B₁₂ is required for normal folate recycling and folate retention within cells, vitamin B₁₂ defi­ciency leads to a failure to retain folate within cells (Hoffbrand & Weir, 2001). Consequently, RBC folate concen­trations fall, despite the presence of normal (or sometimes even elevated) serum folate. Thus, low RBC folate may reflect vitamin B₁₂ defi­ciency as well as folate defi­ciency.

Fasting versus non fasting samples. Samples from fasted individuals, on average, have lower concen­trations of serum (by 10%) and RBC folate (by 5%) compared with samples from nonfasted (< 3h) participants, but the small differences generally indicate that fasting is not essential when assessing the folate status of populations. (Bailey et al., 2015).

Important preanalytical factors. Detailed information on preanalytical factors affecting serum and RBC folate is described in a comprehensive review article on analytical approaches by Pfeiffer et al.,2010. Folate is the least stable of the B vitamins and is susceptible to oxidative degradation during preanalytical sample handling and storage. Careful sample handling and use of anti­oxidants are therefore required to maintain sample integrity. Blood for serum folate analysis should be processed and frozen promptly. Although serum is generally preferred over plasma in most labora­tories, both matrices generally produce comparable results for folate, as long as the sample processing is not delayed. If delays are unavoidable, the sample should be protected from light, kept cool and processed within a few days of collection. Folate in serum and hemolysates (but not in whole blood) can withstand a few short freeze / thawing cycles, particularly if the vials are kept closed as much as possible to minimize the exposure of the sample to air. Folate in serum / plasma degrades rapidly at room temperature, particularly in the presence of EDTA.

Analytical methodologies for measure­ment of serum and RBC folate

Over the past 50y, analytical methods to assess serum and RBC folate concen­trations have been continuously improved; however, they have not yet reached the point where they produce sufficiently comparable results across methods or labora­tories. The within-person variability for serum folate is about twice that for RBC folate (CV of 21.5% and 9.1%, respectively, Bailey et al., 2015).

Critical considerations in folate assess­ment. Assessing folate status is complicated by the large number of folate forms that may be readily interconverted. To overcome this problem, micro­bio­logical assays have been used for many decades due to the ability of some bacteria to grow in the presence of many different forms of folate, i.e. L.rhamnosus responds to all active mono­glutamate forms (see below). Subsequently, assays using competitive protein binding became common because of their simplicity. In more recent years, the use of LC‑MS/MS has emerged. Of note, the emergence of LC‑MS/MS enables investigation of relevant research findings related to folate meta­bolism. These include alterations in the relative proportion of different folate forms in red cells owing to the common MTHFR C677>T poly­morphism, along with an overall lower RBC folate (Molloy et al., 1997; Bagley & Selhub, 1998; Shane et al., 2011), and the presence of free (“unmeta­bolized”) folic acid in the blood as a result of high intakes of folic acid (Pfeiffer et al, 2010).

There are three main method types for measure­ment of serum and RBC folate, each with advantages and disad­van­tages (Bailey et al., 2015).

Microbiolog­ical assay. This is widely considered to be the gold standard assay for serum and RBC total folate because it measures all biolog­ically active forms equally and does not measure folate species that lack vitamin activity. The underlying principle of the micro­biolog­ical assay is that a folate-dependent micro­organism, namely Lactobacillus rhamnosus (formerly called Lactobacillus casei), grows proportionally to the amount of folate present in serum or whole blood and the folate concen­tration can be quantified by measuring the turbidity of the inoculated medium after a 2d incubation. A chloramphenicol-resistant strain of L.rhamnosus is used and the assay is performed using automated micro­titer plate technology.

The key advantages of the micro­biolog­ical assay include its excellent sensitivity, low cost and simple instrumentation (thus suited for low-resource settings), and the fact that it can be used with dried blood spot samples. The high sensitivity is a particular advantage when limited volume is available, such as for samples collected from a fingerstick or dried blood spot. The disad­van­tages are that the micro­biolog­ical assay is relatively laborious unless automated liquid handling is introduced, lengthy assay time with limited throughput and limited linear range (thus requiring dilution of samples). It is also prone to contamination issues and potential inter­ference by the presence of antibiotics or anti­folates. The latter limitation is well recognized, but in practice may not be such an issue, in that analysts in the US reported that < 1% of samples from the population-based NHANES cohort exhibited a pattern of inter­ference due to the presence of anti­biotics or anti­folates (Pfeiffer et al, 2010).

Protein-binding assays. These assays were developed primarily for clinical settings, to enable the diagnosis of folate defi­ciency. Protein-binding assays use the highly specific folate-binding protein to extract folate from the sample. The strengths of this approach include the high sample throughput, quick turn­around time, avail­ability in commercial kit form and minimum operator involvement. The disad­van­tages are that the various folate forms have different affinities to folate-binding protein, the ques­tion­able accuracy when mixtures of folate are present, the limited linear range (thus requiring dilution of samples), matrix effects when sample is diluted, and lot-to-lot variability of commercial kits. Also, although not sensitive to anti­biotics, protein-binding assays are influenced by certain anti­folates such as metho­trexate.

Chromatography-based assays (HPLC-FD, LC‑MS/MS). Chromatography-based methods typically provide information on individual folate forms based on measure­ment of intact folates via HPLC. More recently, LC-tandem mass spec­trometry (LC‑MS/MS) is now the preferred detection for HPLC-based methods in specialized labora­tories. Advantages of chromatography-based approaches are that they measure all folate forms, are highly selective and specific, have good sensitivity and precision, enable in-house control of perform­ance and use of stable-isotope-labeled internal standards to compensate for procedural losses. Ensuring accurate calibration is a big task for chrom­atography-based methods, due to the high number of folate forms and also because of their instability. Disadvan­tages include the high costs and require­ment for expensive instru­men­tation, experienced operator and frequent technical service, along with being a relatively laborious approach unless automated liquid handling is introduced, and the requirement for complex sample extraction / cleanup. Also inter­conversions of folate forms during the assay procedure need to be considered in inter­pretation of data.

Cut-off values and Interpretation of Serum and RBC folate

The measure­ment of total folate provides information on the folate status of the individual, either in the short-term through serum folate, or in the long-term through RBC folate. An historical pers­pec­tive on the use of folate cutoffs is provided in Box 22a.1. However, the inconsistent use of cutoff values over time to assess the proportion of populations with deficient or low folate status led to a certain degree of confusion. More recently, revised cut-off points for folate defi­ciency (serum folate < 10nmol/L and RBC folate < 340nmol/L) have been defined based on the meta­bolic indicator, plasma homo­cys­teine (Selhub et al., 2008), as summarized in Box 22a.1. These cutoff values have been recommended by the WHO for assessing folate status of populations (de Benoist, 2008). Of note, the values were derived from data generated using the micro­biolog­ical assay.

As discussed earlier, the risk of NTD is associated with a maternal folate status that would not conventionally be classed as deficient. Although there is no cutoff value estab­lished by international organizations for folate concen­trations to define NTD risk in populations, it is accepted that the number of NTDs that can be prevented in a population is dependent on maternal folate status. Specifically, RBC folate concen­tration in the mother has been shown to be a sensitive bio­marker of NTD risk. In the one and only prospective study that has been conducted to date, Daly et al. (1995) found that the prevalence of NTD in an Irish population was lowest when maternal RBC folate concen­trations were ≥ 906nmol/L (400ng/mL). Subsequently, data modeled from folic acid intervention studies in China by Crider et al. (2014) confirmed the dose response relationship between RBC folate concentrations and NTD risk as shown in the Irish study (Daly et al., 1995).

22a.2.2 Plasma homo­cys­teine concen­trations

Plasma homo­cys­teine provides a sensitive functional indicator of folate status. On the basis that normal homo­cys­teine meta­bolism requires an adequate supply of folate, plasma homo­cys­teine becomes elevated when folate status is low.

Homo­cys­teine is a four-carbon, thiol-containing amino acid found in human plasma, mostly present in the form of various disulfides, such as homo­cys­teine-cysteine disulfide. About 75% of total homo­cys­teine is bound to protein (mainly albumin), whereas the remainder occurs in nonprotein-bound “free” forms. Total homo­cys­teine is defined as the sum of all homo­cys­teine species in serum or plasma, including free and protein-bound forms. Only a very small portion (1–2%) of plasma homo­cys­teine is present as the thiol, however the relative contribution of the thiol to total homo­cys­teine increases to 10–25% in patients with abnormally elevated plasma homo­cys­teine (Bailey et al., 2015).

Homo­cys­teine is derived from the essential amino acid, methionine. It is meta­bolized in two ways: remethyl­ation to methionine (by methionine synthase) or trans­sulfur­ation to cysta­thionine (by cystathionine beta-synthase; CBS) then to cysteine (Figure 22a.2). These pathways require adequate status of folate and the meta­bolically related B‑vitamins. The remethyl­ation pathway, whereby methionine is synthesized from homo­cys­teine by methionine synthase, is dependent on both folate and vitamin B₁₂ as cofactors, whilst the trans­sulfuration pathway that converts homo­cys­teine to cysteine is catalyzed by two vitamin B₆ dependent enzymes. A fourth B vitamin, ribo­flavin, is required in its cofactor form flavin adenine dinucleo­tide (FAD) for the activity of MTHFR, the enzyme that catalyzes the reduction of 5,10 methyl­eneTHF to 5 methylTHF. Once formed, 5 methylTHF is used by methionine synthase for the vitamin B₁₂-dependent conversion of homo­cys­teine to methionine and the formation of THF. Thus, the concen­tration of homo­cys­teine in plasma / serum is regulated by up to four B‑vitamins: folate, vitamin B₁₂, vitamin B₆ and ribo­flavin.

Apart from providing a functional bio­marker of folate, higher plasma homo­cys­teine is associated with a number of chronic diseases of ageing, including an increased risk of cardio­vascular disease (CVD; Graham et al., 1997), cognitive impairment and dementia (Smith & Refsum, 2016). It remains to be estab­lished, however, whether plasma homo­cys­teine per se a risk factor for these or other diseases. In the case of CVD, this issue is particularly contro­versial. Despite strong and consistent evidence from observational studies over many years, several secondary prevention trials published between 2004 and 2014 failed to demon­strate a benefit of homo­cys­teine-lowering therapy against the recurrence of CVD events in patients with existing disease. The evidence is however stronger for the relation­ship of homo­cys­teine with stroke than heart disease, with good evidence from both population data and randomized trials that folic acid intervention and/or homo­cys­teine-lowering can signif­icantly reduce the risk of stroke, and particularly so in people with no previous history of stroke (McNulty et al., 2017). Of note, although the literature in this area focuses on homo­cys­teine as the CVD risk factor, it is possible that folate and related B‑vitamins have roles in CVD that are independent of their homo­cys­teine-lowering effects. Thus, plasma homo­cys­teine may be linked with CVD as a functional marker of low B‑vitamin status which reliably reflects perturbed one-carbon meta­bolism, rather than being causatively related to CVD per se (McNulty et al., 2017).

Factors affecting homo­cys­teine concen­trations

Folate status. On the basis that normal homo­cys­teine meta­bolism requires an adequate supply of folate, plasma homo­cys­teine is first and foremost affected by folate status. Thus, plasma homo­cys­teine was shown to respond within 3–4wks of folate depletion (increase) and subsequent repletion (decrease) in a controlled meta­bolic study in healthy women (Jacob et al., 1998). Like­wise, in observational studies, plasma homo­cys­teine is invariably found to be inversely associated with folate status, whether measured as serum or RBC folate.

Intervention with folic acid. Plasma homo­cys­teine is highly responsive to intervention with folic acid (the synthetic vitamin form), alone or in combination with vitamin B₁₂, vitamin B₆, ribo­flavin and betaine (or choline). Thus, food fortif­ication with folic acid has marked effects on homo­cys­teine concen­trations. Using population-based data from the US, Pfeiffer et al. (2008) reported a 10% decrease in plasma homo­cys­teine when comparing values pre-fortif­ication (1991–1994) to post-fortif­ication (1999–2004) of food with folic acid, as implemented on a mandatory basis in 1996–1998. Folic acid-fortif­ication on a voluntary basis (i.e. added to food at the manufacturer's discretion) also affects plasma homo­cys­teine. In a convenience sample of nearly 500 healthy adults in Northern Ireland aged 18–92y, who were not taking folic acid supplements, homo­cys­teine concen­trations were lower by 2µmol/L in high consumers compared to non-consumers of fortified foods (providing > 100µg/d and 0µg/d folic acid, respectively) (Hoey et al., 2007).

Folate-related B vitamins. Because normal homo­cys­teine meta­bolism is dependent on four B vitamins, homo­cys­teine concen­trations will be elevated with other B vitamin deficiencies apart from folate, notably vitamin B₁₂ (Allen et al., 2018). Thus whilst plasma homo­cys­teine is primarily a folate bio­marker, once folate status is optimized, a much greater dependency on vitamin B₁₂ emerges (Quinlivan et al., 2002). Like­wise, in population groups who consume folic acid-fortified foods or folic acid supplements, homo­cys­teine is considered to be a more reliable bio­marker of vitamin B₁₂ than folate (Refsum et al., 2004).

Disease and lifestyle. Apart from inversely reflecting folate and related B vitamin status, plasma homo­cys­teine is found to be elevated in patients with impaired renal function (Yetley et al., 2011; Allen et al., 2018) and with certain drug treat­ments (Refsum et al., 2004). Alcohol intake, coffee-drinking and smoking are also associated with higher plasma homo­cys­teine.

Age, sex and life­cycle stage. Homo­cys­teine concen­trations increase throughout life and are higher in males compared to females at all ages. Although plasma homo­cys­teine decreases in preg­nancy (by about 50%), concen­trations are reported to normalize within a few days post­partum (Murphy et al., 2004).

Ethnic and genetic effects. Plasma homo­cys­teine may also differ among ethnic groups, but these differences appear to be less important than the influence of B vitamin status. Notably, the most common genetic cause of elevated homo­cys­teine in the general population is homozygosity for the MTHFR C677>T poly­morphism (affecting an estimated 10% of people worldwide but much higher in some populations, including Mexico and Northern China). This folate gene variant also contributes to a higher risk of blood pressure throughout the lifecycle, particularly when combined with deficient status of ribo­flavin, which is required as a cofactor for MTHFR, a key folate meta­bolizing enzyme (Psara et al., 2020; Ward et al., 2020). Notably, both phenotypes associated with this common poly­morphism (high homo­cys­teine and high blood pressure) appear to be modifiable with better ribo­flavin status (McNulty et al., 2020). Most other genetic poly­morphisms in enzymes related to one-carbon meta­bolism have little effect on homo­cys­teine concen­trations (Refsum et al., 2004).

Inborn errors of meta­bolism. Homocystinuria refers to rare inborn errors of meta­bolism which lead to severely increased homo­cys­teine concen­trations in plasma, usually > 100µmol/L, along with large amounts excreted in urine. The most common cause of homo­cystin­uria is defi­ciency of the enzyme cysta­thionine β‑synthase (CBS). CBS defi­ciency, an autosomal recessive condition, has a reported world­wide birth prevalence of 1 in 344,000, while that in Ireland the frequency is much higher, at 1 in 65,000, based on newborn screening and cases detected clinically. These patients have a high risk of premature, frequently fatal, throm­boembolic events. Early diagnosis and treatment with pyridoxine and/or folic acid and betaine, preferably from infancy, can however prevent CVD events and most of the clinical symptoms (Refsum et al., 2004).

Analytical methodologies for measure­ment of plasma homo­cys­teine

Plasma total homo­cys­teine is a very stable analyte as long as the plasma is separated from the red blood cells within 1hr of blood collection (or within 8 hours if the whole blood is kept on ice). As described in detail by Refsum et al. (2004), various method types are available for homo­cys­teine determination. These range from fully automated commercial kits (immunoassay or enzymatic methods) to chrom­atographic assays with mass spec­trometry detection, overall providing comparable results and good assay performance. All methods require the reduction of the disulfide bonds to allow measure­ment of total homo­cys­teine. The reported within-person variability CV for plasma homo­cys­teine is 12.2% (Refsum et al., 2004).

Choice of method. Because the measure­ment of plasma homo­cys­teine produces comparable results across different method types, the choice of method is mainly dependent on available instrumentation and technical expertise. Bailey et al. (2015) put forward the following recommendations to help in making this decision:

• Using commercial kits (either immunoassay or enzymatic assay) on a fully-automated clinical analyzer will provide the highest throughput and quickest turnaround time with the least effort. The main disad­van­tage is the relatively high reagent costs, which can make the measure­ments quite expensive, particularly for a large number of samples.
• If an HPLC system with fluorescence detector is available, setting up a manual assay may be the least expensive approach, particularly in settings with low labor costs. The disad­van­tage of the manual HPLC assay is that the number of samples that can be run is quite limited (typically to about 50–70 unknown samples per analytical run).
• If a labora­tory has access to an existing LC‑MS/MS system and the required technical knowhow, plasma homo­cys­teine can be measured in a high throughput semi-automated manner, with quite low reagent costs. Due to the high initial cost, however, an LC‑MS/MS system is economical only if a large number of samples are measured regularly.
• GC-MS or GC-MS/MS have also been used in high throughput research settings, particularly when the co-determination of methyl­malonic acid and methionine is of interest.
• GC-based methods provide better precision, higher resolution, and longer column life compared to LC-based methods and are not subject to ion suppression issues, which can be a problem in LC‑MS/MS methods.

Cut-off values and Interpretation

In one of the most comprehensive reviews of all relevant aspects of homo­cys­teine measure­ment, it is proposed that reference values for homo­cys­teine are estab­lished for different populations to account for the influence of both nonmodifiable and modifiable factors (Refsum et al., 2004; Table 22a.1).

Table 22a.1. Upper reference limits for plasma homocysteine (µmol/L) in populations. Adapted from: Refsum et al., (2004) with Data drawn from the US, Norway, the UK and Israel and from 9 European countries participating in the European Concerted Action Project (Graham et al., 1997)
¹Individuals eating folic acid-fortified food or taking folic acid-containing supplements.

Population sub-group	Folic acid Supplemented1	Non- supplemented
Pregnancy	> 8	> 10
Children < 15y	> 8	> 10
Adults 15–65y	> 12	> 15
Adults > 65y	> 16	> 20

The following recommendations were put forward for setting reference ranges:

• Each labora­tory should establish reference limits for its region.
• The reference population should exclude individuals with folate or cobalamin defi­ciency or increased creatinine, and patients with diseases or who are taking drugs that increase homo­cys­teine concen­trations.
• Separate reference limits for children, adults, the elderly, and preg­nant women should be used (Table 22a.1).
• In populations consuming folic acid supplements or folic acid-fortified foods, the upper reference limit is usually 20–25% lower than in non­fortified populations (Table 22a.1).
• The upper homo­cys­teine reference limit should be inter­preted with age, sex, and other determin­ants in mind, particularly prevailing food fortif­ication practices.

22a.2.3 Other bio­markers of folate status

Neutro­phil lobe count

Usually, neutro­phils have three or four segments, but in megalo­blastic anemia (owing to folate or vitamin B₁₂ defi­ciency), this number increases. Neutro­phil hyper­segmentation can be evaluated in smears of peripheral blood or in white blood cells obtained from the buffy coat (i.e. interface between the serum / plasma and the sedimented red cells), and may be the earliest morphological change to appear in the blood in folate and B₁₂ defi­ciency. Neutro­phil hyper­segmentation however is not specific to folate or B₁₂ defi­ciency as it also occurs in other conditions including uremia, myelo­pro­liferative disorders, myelo­fibrosis and as a congenital lesion in approximately 1% of the population, even when the status of folate and vitamin B₁₂ is adequate. The observed greater incidence of neutro­phil hyper­segmentation in iron defi­ciency anemia concurrent with folate and/or vitamin B₁₂ defi­ciency implies interactive effects on the erythro­poietic process (Westerman et al., 1999; Metz, 2008).

Serum folic acid

Unmeta­bolized (or “free”) folic acid can be present in varying concen­trations in plasma and is typically associated with higher folic acid intake via fortified foods, supplements, or a combination of both (Bailey et al., 2010). Whilst greater concen­trations of unmeta­bolized folic acid are generally associated with higher serum folate concen­trations, there is large variation in reported values and no clear dose response relation­ship exists between folic acid intake and unmeta­bolized folic acid in plasma (Pfeiffer et al., 2004; Bailey et al., 2010; Obeid et al., 2011). Thus, cut-off values or desirable ranges have yet to be identified. One randomized trial addressed concerns relating to the potential for folic acid to exert adverse biological effects in pregnancy, and demonstrated that folic acid supplements taken at recommended levels throughout pregnancy do not lead to increased circulating unmetabolized folic acid in mothers or their babies (Pentieva et al., 2016). However, the biological impacts and the significance for health in general of unmetabolized folic acid in blood remain to be elucidated.

Genomic bio­markers of folate status

DNA cytosine methy­lation. Apart from its role in DNA synthesis, folate plays an important role in DNA meta­bolism as it is required for the synthesis of methionine and thus SAM, which in turn is required as a methyl donor for the maintenance of cytosine methy­lation, essential for silencing of genes or structural integrity of specific regions of chrom­osomes (Fenech, 2012). When SAM is depleted, the maintenance of DNA methy­lation may become compromised, leading to hypo­methy­lation of cytosine and structural changes in chrom­atin. Some studies suggest that global DNA methy­lation status is reduced when folate is deficient (Kim, 2005; Crider et al., 2012), but findings in this regard are somewhat inconsistent.

Uracil misincorporation into DNA. Measurement of uracil content in DNA may also provide a bio­marker of folate status, on the basis that adequate amounts of folate as 5,10‑methyl­eneTHF are required by thymidylate synthase to convert deoxy­uridine mono­phosphate (dUMP) to thymidylate mono­phosphate (dTMP) in pyrimidine synthesis and thus DNA biosynthesis. If 5,10‑methyl­eneTHF is limiting as a substrate of thymidylate synthase, dUMP accumulates and it becomes more probable that uracil is incorporated into DNA instead of thymidine during DNA synthesis (Stover, 2009; Fenech, 2012).

Micronuclei. Micronuclei have the same morphological features as normal nuclei but are much smaller. Excessive uracil incorporation into DNA and hypo­methy­lation of DNA can lead to the formation of micro­nuclei from chrom­osome fragments which can be measured in lympho­cytes or in erythro­cytes (Blount et al., 1997; Fenech, 2012). Observational studies show that micro­nuclei in lympho­cytes or erythro­cytes are inversely associated with dietary folate intake and RBC folate (Blount et al., 1997), whilst intervention studies have reported a signif­icant reduction in micro­nuclei frequency in lympho­cytes with folic acid supple­men­tation (Fenech et al., 1997). It is now possible to score micro­nuclei automatically and reliably using a wide range of image cytometry platforms making this technique amenable to mass screening. Given its sensitivity to folate defi­ciency, micro­nuclei measure­ment, in combination with uracil and DNA methy­lation measure­ments, can potentially provide a reliable assess­ment of genome pathology resulting from deficient folate status (Bailey et al., 2015).

22a.2.4 Assessment of folate status in children

Most studies assessing folate status have focused on adults. Far fewer studies provide folate bio­marker data for children, with notable exceptions being published reports using data from population-based surveys conducted in the United States (Pfeiffer et al., 2012) and the United Kingdom (Kerr et al., 2009;. Figure 22a.3) The latter report proposes normal ranges for folate bio­markers for use in clinical pediatric settings

Population-based data in British and American children, both show progressive declines in folate (and corresponding, increases in homo­cys­teine) concen­trations with age from childhood to adolescence. Consistent with these reports, are the findings from convenience cohorts of Belgian, Dutch, and Greek children, also showing age-related decreases in folate concen­trations in childhood (De Laet et al., 1999; van Beynum et al., 2005; Papandreou et al., 2006). Of note, where dietary intakes were also measured, the data showed that folate intakes generally compared favorably with dietary reference values across all age groups and were not lower in the older children (Papandreou et al., 2006; Kerr et al., 2009; Pfeiffer et al., 2012) The explanation for the decline in folate status bio­markers with age in children, despite no corresponding decline in dietary folate intakes, is not entirely clear but likely reflects the higher folate requirements of older children related to increased meta­bolic demands for growth from childhood to adolescence (Bailey et al., 2015).

22a.3 Conclusions

Folate is required for one-carbon meta­bolism and thus plays a critical role in essential biolog­ical processes, including amino acid meta­bolism, DNA synthesis and repair and methy­lation reactions. Low folate status leads to adverse health effects throughout the lifecycle, even if not severe enough to cause mega­loblastic anemia, the clinical manifestation of folate defi­ciency. Thus, accurate assess­ment of folate status is essential. Serum and RBC folate concen­trations are sensitive bio­markers used widely to assess folate status. Serum folate is the earliest indicator of altered folate exposure and reflects recent dietary intake, whilst RBC folate reflects tissue folate stores and is a better indicator of folate intake over the previous 3–4 months when circulating folate is incorporated into the maturating red cells. On the basis that normal homo­cys­teine meta­bolism requires an adequate supply of folate, the measure­ment of plasma homo­cys­teine provides a sensitive functional indicator that will be elevated with deficient or low folate status, but it is not a specific folate bio­marker as it is influenced by other nutrient (most notably vitamin B₁₂ defi­ciency) and non-nutrient factors including renal function. Emerging genomic bio­markers may provide a reliable assess­ment of genome pathology resulting from deficient status to folate, to add to the core bio­markers of status, serum folate, RBC folate, and plasma homo­cys­teine.

Acknowledgements

Some of the research described in this chapter (22a) was supported by grants from: the UK Food Standard Agency (3 consecutive awards; 1998–2008); DSM Nutritional Products, Switzerland; the Health and Social Care Public Health Agency Northern Ireland (Enabling Research Awards scheme STL/5043/14); the European JPI ERA-HDHL Nutrition & the Epigenome scheme, funded by UKRI (BB/S020330/1) incorporating the Biotechnology and Biological Sciences Research Council (BBSRC) and the Medical Research Council. None of these entities was involved in the writing of this chapter.