Roles of riboflavin in human health

Riboflavin is a water-soluble B-vitamin, which acts as a precursor for the coenzymes FMN and FAD, the metabolically active vitamin forms⁽ Reference Pitkin ^{1 )}. These coenzymes (FAD and FMN) participate in intermediary metabolism and catalyse numerous oxidation–reduction reactions, playing a fundamental role in the metabolism of energy, certain drugs and toxins and in antioxidant protection⁽ Reference McCormick and Shils ^{8 ,} Reference Powers ^{9 )}. Furthermore, riboflavin interacts with a number of other nutrients, including metabolically linked B-vitamins and iron. Early animal studies have linked riboflavin deficiency with impaired iron absorption, increased intestinal loss of iron, and/or impaired iron utilisation for the synthesis of Hb⁽ Reference Powers, Weaver and Austin ^{10 )}. More recently, supplementation with riboflavin has been shown to enhance circulating Hb levels in human subjects⁽ Reference Powers, Hill and Mushtaq ^{11 )} furthermore, correcting riboflavin deficiency in individuals who were both riboflavin and iron deficient improved the response of iron deficiency anaemia to iron therapy⁽ Reference Powers ^{12 )}. Irrefutable evidence has shown the metabolism of other B-vitamins is dependent on riboflavin coenzymes. Riboflavin is involved in vitamin B₆ metabolism; the enzyme pyridoxine-phosphate oxidase requires FMN for the conversion of pyridoxine phosphate to its coenzyme form pyridoxal-5 phosphate⁽ Reference McCormick ^{13 )}. Historical evidence from animal studies reported that pyridoxine-phosphate oxidase activity is sensitive to changes in dietary riboflavin intake and thus riboflavin deficiency may alter pyridoxal 5′ phosphate activity⁽ Reference Rasmussen, Barsa and McCormick ^{14 )}. Research from our centre conducted a number of years ago confirmed the interrelationship between riboflavin and vitamin B₆ in human subjects and showed that riboflavin supplementation of older adults not only improved biomarker status of riboflavin, but also enhanced blood pyridoxal-5 phosphate (vitamin B₆) concentrations⁽ Reference Madigan, Tracey and McNulty ^{15 )}. Niacin synthesis is also reliant on the FAD-dependent enzyme kynurenine mono-oxygenase, which is required for the synthesis of the coenzymes NAD and nicotinamide adenine dinucleotide phosphate from tryptophan. Riboflavin deficiency decreases the conversion of tryptophan to NAD and nicotinamide adenine dinucleotide phosphate resulting in niacin deficiency⁽ Reference McCormick and Shils ^{8 )}. Work from our centre and others have demonstrated the key role riboflavin plays in C₁ metabolism via its role as a co-factor for the MTHFR enzyme.

Riboflavin absorption and transport

Dietary riboflavin occurs mainly in the form of FAD and smaller amounts occur as FMN or as free riboflavin⁽ Reference Powers ^{16 )}. Unlike free riboflavin, FAD and FMN must be hydrolysed in the intestinal lumen to yield free riboflavin prior to absorption. Animal studies have shown that the uptake of dietary riboflavin from the intestine is increased in riboflavin deficiency⁽ Reference Said and Mohammadkhani ^{17 )} and urinary excretion was found to increase linearly with increasing dietary intakes in individuals with optimal status⁽ Reference Sauberlich, Skala and Dowdy ^{18 )}. The transport of flavins in blood is by loose binding to albumin and tight binding to a number of immunoglobulins in serum particularly IgA, IgG and IgM⁽ Reference McCormick ^{19 )}. A number of physiological factors have been reported to influence the rate of intestinal absorption of riboflavin. Diets high in psyllium gum decrease the rate of intestinal absorption, whereas bile salts increase absorption⁽ Reference Rivlin, Wyngaarden, Smith, Bennett and Plum ^{20 )}. Alcohol is reported to interfere with the digestion of food flavins into riboflavin and the intestinal absorption of riboflavin⁽ Reference Pinto, Huang and Rivlin ^{21 )}. Notably, concentrations of riboflavin synthesised by bacterial metabolism in the human colon may be more than 6-fold higher than dietary intakes⁽ Reference Yuasa, Hirobe and Tomei ^{22 )}.

Riboflavin requirements and sources

Worldwide dietary recommendations for riboflavin range from 1·1 to 1·6 mg/d for adults, an increment of 0·3 mg/d is recommended during pregnancy to cover the increased tissue synthesis for fetal and maternal development and an additional 0·4–0·5 mg/d during lactation^{( 23 – 25 )}. Clinical signs of deficiency in human subjects appear at intakes <0·5–0·6 mg/d and urinary excretion of riboflavin is seen at intake levels of approximately 1 mg/d⁽ Reference Pitkin ^{1 )}. Riboflavin is found in a wide variety of food but yeast extract and offal products especially those based on liver are the only rich sources, containing more than 2 mg/100 g^{( 26 )}. The latest National Diet and Nutrition Survey reported that milk and milk products, meat, and fortified breakfast cereal make the greatest contribution to riboflavin intake in a British diet⁽ Reference Bates, Lennox and Bates ^{27 )} (Table 1). Unlike any European country, the USA has a mandatory riboflavin enrichment policy to replace the riboflavin lost from starch during milling (0·40 mg for wheat starch)⁽ Reference Allen, de Benoist and Dary ^{28 )}. Similarly, a mandatory enrichment of starch with riboflavin (4 mg/kg for starch) is in place in Canada. No toxic effects have been reported in relation to intakes of riboflavin at doses higher than dietary recommendations⁽ Reference McCormick, Shils, Olsen and Shike ^{29 –} Reference Cooperman, Lopez and Machlin ^{32 )}, the absorption of riboflavin appears to be limited to approximately 30 mg at any one time⁽ Reference McCormick ^{13 )} with apparently little or no absorption at higher doses⁽ Reference Zempleni, Galloway and McCormick ^{33 )}. Furthermore, the Institute of Medicine in the USA did not establish a tolerable upper intake level for riboflavin when the RDA was revised in 1998⁽ Reference Pitkin ^{1 )}. A harmless side-effect of high-dose riboflavin intake is the increased intensity of urine colour to bright yellow.

Table 1.

Food sources of riboflavin in the UK (NDNS)⁽ Reference Bates, Lennox and Bates ^{27 )}

NDNS, National Diet and Nutrition Survey (2008–2009 to 2011–2012).

Riboflavin deficiency

Isolated clinical riboflavin deficiency does not have unique or characteristic physical features. The classical symptoms, angular stomatitis, cheilosis and glossitis, are not specific to riboflavin deficiency and may be due to other vitamin deficiencies. Furthermore, after 3–8 months of inadequate riboflavin intake other symptoms are reported to appear including magenta tongue, seborrheic dermatitis, vascularisation of the cornea and normochromic normocytic anaemia⁽ Reference Pitkin ^{1 ,} Reference McCormick and Shils ^{8 )}. Other than dietary inadequacy certain endocrine abnormalities i.e. adrenal and thyroid hormone insufficiency⁽ Reference Rivlin, Wyngaarden, Smith, Bennett and Plum ^{20 ,} Reference Cimino ^{34 )} and certain drugs can inhibit the conversion of riboflavin into its active coenzyme derivatives⁽ Reference Apeland, Mansoor and Pentieva ^{35 ,} Reference Dutta, Pinto and Rivlin ^{36 )}.

Newborn infants are at increased risk of riboflavin deficiency when maternal status is poor during pregnancy⁽ Reference Bates, Prentice and Paul ^{37 )} or as a result of phototherapy treatment for hyperbilirubinaemia⁽ Reference Tan, Chow and Karim ^{38 )}. Pregnant women with riboflavin deficiency have been reported to be 4·7 times more likely to develop preeclampsia compared to those with adequate riboflavin status⁽ Reference Wacker, Fruhauf and Schulz ^{39 )}. It has been suggested this may be associated with mitochondrial function, oxidative stress and blood vessel dilation⁽ Reference Wacker, Fruhauf and Schulz ^{39 )}. Furthermore, a common polymorphism (C677T) in the gene encoding the FAD-dependent enzyme MTHFR has been associated with an increased risk of preclampsia⁽ Reference Wu, Yang and Tang ^{40 )}. A number of studies have reported that riboflavin requirements are higher with increased physical activity levels⁽ Reference Belko, Obarzanek and Roach ^{41 ,} Reference Soares, Satyanarayana and Bamji ^{42 )}, excessive alcohol consumption and smoking⁽ Reference Pinto, Huang and Rivlin ^{21 ,} Reference Benton, Haller and Fordy ^{43 )}.

Novel role of MTHFR genotype and blood pressure

Globally, hypertension accounts for 16·5 % of deaths each year (9·4 million); an estimated 45 % of deaths due to heart disease and 51 % of deaths due to stroke are a result of hypertension⁽ Reference Lim, Vos and Flaxman ^{79 )}. High blood pressure (BP), even within the normal range substantially increases the risk of CVD and death, while a lowering of systolic BP by as little as 2 mmHg can decrease CVD risk by as much as 10 %⁽ Reference Lewington, Clarke and Qizilbash ^{80 )}. Hypertension is a polygenic disease that occurs as a result of a complex interaction of diverse environmental conditions and genetic factors. Risk factors include high dietary sodium intake, excess weight, excessive alcohol consumption and lack of physical activity⁽ Reference Whelton, He and Appel ^{81 )}. Notably, twin studies have reported the heritability of BP variation to be as much as 50 %⁽ Reference Luft ^{82 ,} Reference Mongeau ^{83 )}. In the past decade, genome-wide association studies have identified several genetic loci associated with BP variation in European populations, including one near the gene encoding the folate metabolising enzyme MTHFR⁽ Reference Tomaszewski, Debiec and Braund ^{84 –} Reference Newton-Cheh, Johnson and Gateva ^{86 )}. This finding was also replicated in non-European cohorts including Chinese, Japanese, Indian and US populations⁽ Reference Takeuchi, Isono and Katsuya ^{87 –} Reference Flister, Tsaih and O'Meara ^{90 )}.

Generally consistent with these findings from genome-wide association studies, there is a growing body of evidence from observational studies to support a specific association between the 677C → T polymorphism in the MTHFR gene and BP variation⁽ Reference Heux, Morin and Lea ^{91 –} Reference Bayramoglu, Kucuk and Guler ^{94 )}. As the evidence has accumulated numerous meta-analysis were conducted, all reporting significant associations of the MTHFR 677C → T polymorphism with hypertension (Table 2). Although there is strong evidence linking this polymorphism with hypertension, a number of observational studies have reported conflicting or inconclusive results. Many of the studies that failed to detect significant associations used small sample sizes or suffered possible selection bias⁽ Reference Nishio, Lee and Fujii ^{93 ,} Reference Benes, Kankova and Muzik ^{95 –} Reference Fowdar, Lason and Szvetko ^{99 )}. Some studies have reported a gender specific association; one large population study in >3000 Japanese individuals reported that the MTHFR 677C → T polymorphism was associated with a 42 % increased risk of hypertension in women but not in men⁽ Reference Inamoto, Katsuya and Kokubo ^{96 )}. In contrast, in a Spanish cohort, the TT genotype was found to be a significant predictor of hypertension in men but not in women, however, only 26 % of the sample were females⁽ Reference Rodriguez-Esparragon, Hernandez-Perera and Rodriguez-Perez ^{100 )}.

Table 2.

Meta-analyses of association of C677T polymorphism in MTHFR with hypertension*

MTHFR, methylenetetrahydrofolate.

* A number of these meta-analyses have considered hypertension in pregnancy, only results for hypertension have been considered in this review.

† Odds ratio refers to MTHFR CC v. TT genotypes.

Recent work conducted at our centre has considered the BP-lowering effect of riboflavin supplementation in individuals with the MTHFR 677TT genotype. The first of these trials was conducted in premature CVD patients (mean age 54 years) and demonstrated that riboflavin (1·6 mg/d for 16 weeks) decreased systolic BP (−13·2 (sd 15·0) mmHg; P ≦ 0·02) and diastolic BP (−7·5 (sd 12·0) mmHg; P = 0·02) specifically in individuals with the TT genotype, while no BP response was observed in those with CC or CT genotype⁽ Reference Horigan, McNulty and Ward ^{5 )}. These findings were later confirmed in a 4-year follow-up cross-over design study, which demonstrated in the same cohort (mean age 59 years) that riboflavin (at the same dose and duration of intervention) significantly lowered systolic BP (−9·2 (sd 12·8) mmHg; P = 0·001) and diastolic (−6·0 (sd 9·9) mmHg; P = 0·003) BP specifically in the TT genotype group⁽ Reference Wilson, Ward and McNulty ^{6 )}. These findings were subsequently confirmed in hypertensive patients without overt CVD aged 70 years (5·6 (sd 2·6) mm Hg lowering in systolic BP; P = 0·033)⁽ Reference Wilson, McNulty and Ward ^{7 )}. The extent of response to riboflavin supplementation observed in these trials appears to lessen with increasing age. Based on the available evidence⁽ Reference Horigan, McNulty and Ward ^{5 –} Reference Wilson, McNulty and Ward ^{7 )} and in agreement with preliminary findings from a large population-based study⁽ Reference Reilly, McNulty and Ward ^{101 )}, it appears that age is a significant factor in relation to the BP phenotype and its responsiveness to riboflavin and should be considered in future studies. To date, a low-dose supplementation level (1·6 mg/d) of riboflavin has been used and the effect of higher doses is not known. Thus, it remains possible that greater BP-lowering could be achieved with a larger dose of riboflavin. Of note, in all of the aforementioned trials the BP-lowering effect of riboflavin was shown to be independent of the number and type of antihypertensive drugs being currently administered.

Mechanism of MTHFR C677T polymorphism, riboflavin and blood pressure

The exact mechanism by which the MTHFR C677T polymorphism affects BP (and riboflavin modulates the relationship) has not been clearly identified; however, there are a number of plausible explanations which could explain these effects. In two separate studies involving patients undergoing coronary artery bypass graft surgery, it was identified that those with the MTHFR 677TT genotype had reduced vascular concentrations of 5-methyltetrahydrofolate which in turn were associated with deregulation of nitric oxide (NO); a potent vasodilator known to play a key role in BP⁽ Reference Antoniades, Shirodaria and Warrick ^{102 ,} Reference Antoniades, Shirodaria and Leeson ^{103 )}. This group and others have not considered the role of riboflavin; however, riboflavin supplementation, could in theory stabilise the variant MTHFR enzyme and restore 5-methyltetrahydrofolate concentrations in vascular cells, thereby improving NO bioavailability, which could in turn improve endothelial function and lower BP in individuals with the TT genotype. It is also possible that this novel gene–nutrient interaction may be a result of an imbalance of non-methylated folate derivatives in the endothelial cells in individuals with the TT genotype which in turn could reduce endothelial NO synthase coupling. In individuals with the TT genotype an accumulation of formylated tetrahydrofolates has been detected in erythrocytes, while only 5-methyltetrahydrofolate was found in the erythrocytes of individuals with the CC genotype⁽ Reference Bagley and Selhub ^{104 )}. It has been suggested that an accumulation of 10-formyl tetrahydrofolate in endothelial cells may affect folate metabolism and in turn affect endothelial NO synthase activity⁽ Reference Strain, Hughes and McNulty ^{105 )}. Therefore, riboflavin may enhance endothelial NO synthase activity by correcting the imbalance in methylated v. non-methylated tetrahydrofolate in those with the TT genotype or by increasing 5-methyltetrahydrofolate and thus decreasing BP irrespective of BP-lowering drugs⁽ Reference Strain, Hughes and McNulty ^{105 )}. Further work is required to investigate mechanisms linking this polymorphism with BP and the potential for riboflavin to provide a targeted option to treat elevated BP in this genotype group.

Conclusion and future work

There is emerging evidence that sub-optimal riboflavin status is a problem not confined to developing countries, but also evident in the developed world. Apart from the widely recognised roles of riboflavin in human health, a novel and important role of riboflavin in modulating BP specifically in individuals with the MTHFR 677TT genotype is emerging. Optimal riboflavin status may therefore be particularly important in maintaining health for the 10 % of individuals worldwide (and up to 30 % in some populations) who share this genetic characteristic and are thus at increased risk of developing hypertension. Riboflavin potentially offers a personalised approach to the prevention and treatment of hypertension in these genetically at risk individuals. Further studies are however required to further investigate the BP-lowering effect of riboflavin in different populations and in response to doses higher than 1·6 mg/d. Furthermore, work focusing on the translation of this research to health professionals and patients is also required.