Revised D-A-CH Reference Values for the Intake of Vitamin B₆

The D-A-CH “reference values for nutrient intake” [1] are jointly issued by the Nutrition Societies of Germany, Austria, and Switzerland (the abbreviation D-A-CH arises from the initial letters of the common country identification for the countries Germany [D], Austria [A], and Switzerland [CH]). Reference value is a collective term for recommended intake (RI), estimated values, and guiding values. The starting point for the derivation of reference values is the determination of the average requirement (AR). The experimentally ascertained AR assumes to meet the needs of the daily nutrient intake of 50% of a defined group of people. For the derivation of the RI, 20–30% is added to the AR what corresponds to 2 standard deviations or a variation coefficient of 10–15%. Thus, a RI value, according to its definition, meets the requirement of nearly any person (approximately 98%) of sex- and age-stratified population groups. Estimated values are given when human requirements cannot be determined with desirable accuracy. Guiding values are given as orientation [1, 2].

Reference values for nutrient intake are amounts that are assumed to

• protect nearly all healthy individuals in a population from deficiency-related conditions,

• ensure optimal physiological and psychological performance, and

• create a certain body reserve [1].

The reference values should be revised regularly. Since 2012, the D-A-CH nutrition societies have published revised reference values for the intake of several nutrients [3-13] but not yet for vitamin B₆. The last update of the reference values for vitamin B₆ given by the D-A-CH nutrition societies dates to a report in 2000 [14]. In summer 2019, the revised reference values for vitamin B₆ intake were published in German. This paper provides a summary of this work.

Vitamin B₆ is one of the water-soluble vitamins [15]. The term refers to a group of 2-methyl-3-hydroxy-5-hydroxymethyl pyridine compounds with pyridoxine activity, including pyridoxine (PN), pyridoxamine (PM), and pyridoxal (PL) as well as their phosphoric acid esters pyridoxine 5′-phosphate (PNP), pyridoxamine 5′-phosphate (PMP), and pyridoxal 5′-phosphate (PLP) [16].

Physiologically, PLP and PMP function as cofactors for enzymes of the amino acid metabolism [16]. In addition, vitamin B₆ plays a role in metabolism of carbohydrates [17], conversion of niacin to tryptophan [18], and synthesis of sphingolipids [19], 5-aminolevulinic acid (a precursor of haemoglobin) [20], and neurotransmitters [21]. Together with riboflavin, folate, and cobalamin, vitamin B₆ regulates homocysteine metabolism [22]. In this context, PLP is required for the conversion of 5,10-methylenetetrahydrofolate from tetrahydrofolate and thus for the remethylation of homocysteine to methionine [22]. Within the transsulfuration pathway, vitamin B₆ is involved in the degradation of homocysteine to cysteine [22].

The vitamin B₆ content in foods was found to be affected by freezing, cooking, and further processing, albeit the extent of the losses varied depending on the assessed compound [23]. The absorption rate of vitamin B₆ depends on the diet. Vitamin B₆ from animal-based foods has a higher bioavailability than vitamin B₆ from plant-based foods [24] due to the lower bioavailability of pyridoxine-5′-β-d-glucoside (PNG) found in plant-based foods compared to PN [24-26]. The estimated bioavailability of free PN is ≥95% [27, 28], while the average bioavailability of PNG is 50–58% [25, 26]. Depending on the kind of food, the PNG proportion of the total vitamin B₆ content ranges between 0 and 82% [24]. With a mixed diet containing 2.3 mg of vitamin B₆ per day, the average absorption rate of vitamin B₆ is 75%, based on measurements of plasma PLP concentration and 24-h urinary vitamin B₆ excretion in men [27].

The average vitamin B₆ pool in the human body is estimated to be ∼1,000 μmol (167 mg), whereupon up to 80% of the vitamin B₆ is located in the muscle [29, 30]. The majority of vitamin B₆ in the muscle is found as PLP bound to glycogen phosphorylase [21, 29]. The vitamin B₆ pool in muscle is relatively stable, even at a low-vitamin B₆ diet for 6 weeks [31]. In plasma, vitamin B₆ exists primarily as PLP [32] bound to albumin [33]. In erythrocytes, PLP binds to haemoglobin [16]. Free PLP can be dephosphorylated by alkaline phosphatase to PL, which is capable of crossing biological membranes [16].

There is a sigmoidal relationship between vitamin B₆ intake and plasma PLP concentration [27]. Plasma PLP concentration reaches a new equilibrium ∼10 days after PN supplementation [34] or about 14 days after changing dietary vitamin B₆ intake [27]. The half-life of vitamin B₆ is 2–6 weeks; around 2–3% of the vitamin B₆ pool are lost per day [35]. 4-Pyridoic acid (4-PA) is the major catabolic metabolite with a proportion of up to 85% of the total vitamin B₆ excretion via urine [34].

Vitamin B₆ status can be determined directly by biomarkers, for example, concentrations of PL, PMP, or PLP in plasma or erythrocytes, and total vitamin B₆ or 4-PA concentrations in urine [16]. Currently, plasma PLP concentration is considered to be the best status parameter [28]. It correlates with both vitamin B₆ intake [36-38] and vitamin B₆ concentration in tissue [34]. A deficient vitamin B₆ status is assumed for PLP concentrations <20 nmol/L. PLP concentrations between 20 and 30 nmol/L indicate a marginal status and PLP concentrations >30 nmol/L an adequate supply [39]. In intervention studies in healthy subjects who received a low-vitamin B₆ diet (<0.5 mg vitamin B₆ per day), changes in the amino acid profile and in the metabolism of one-carbon units and tryptophan occurred at PLP concentrations <30 nmol/L [39, 40]. It has to be considered that inflammation [41-43], a low albumin concentration, and a high alkaline phosphatase activity may have a negative impact on PLP concentration [16]. Lifestyle factors like smoking and alcohol intake are considered to be influencing factors as well [43-45]. Current and former smokers show frequently lower PLP concentrations than subjects who had never smoked what might be a result of oxidative stress caused by smoking [43]. Some studies reported a positive association between vitamin B₆ status parameters and alcohol consumption [46, 47]. This observation might be explained by a higher vitamin B₆ intake due to the consumption of alcoholic beverages or changes in transsulfuration and methylation processes [46]. In contrast, chronic alcohol abuse is associated with lower PLP concentrations [44], probably due to an impairment of liver function [45].

Since vitamin B₆ is important for amino acid metabolism, it was proposed that the requirement for vitamin B₆ is linked to protein intake [48]. However, due to inconsistent results, different study designs, and varying protein intake levels [48-52], no influence of protein intake on vitamin B₆ requirements of subjects following a usual mixed diet is assumed.

The reference values for vitamin B₆ intake in adults under 65 years of age were derived based on the intake levels that are required for a plasma PLP concentration of ≥30 nmol/L. According to the results of intervention studies, daily intakes of 1.3–1.5 mg vitamin B₆ are sufficient to achieve a PLP concentration of ≥30 nmol/L in the majority of women, even at a daily protein intake of 1.2 or 1.55 g/kg body weight reported in some of these intervention trials [37, 53-55]. According to the combined analysis of 5 balance studies [37, 49, 53-55], the average vitamin B₆ requirement to ensure a plasma PLP concentration of ≥30 nmol/L is 1.2 mg/day, considering the higher bioavailability of synthetic PN [37]. Assuming a coefficient of variation of 10% (addition of 20%), the RI for women aged 19 to under 65 years is 1.4 mg/day (Table 1).

As there are no convincing balance studies in men, their requirement was extrapolated from the derived requirement in women. Assuming that vitamin B₆ requirement depends on the metabolically active body mass [28], and taking into account an allometric exponent, the AR in men is 1.4 mg/day. Allometric scaling is based on the association between body mass and metabolism. According to Kleiber’s law, the metabolism scaled with a potency of 0.75 of body mass [56, 57]. Due to the dependence of B vitamins on the metabolic body mass, an allometric exponent of 0.75 is included in the calculation. Considering a coefficient of variation of 10% (addition of 20%), the RI for men aged 19 to under 65 years is 1.6 mg/day (Table 2).

Several cross-sectional studies showed an inverse association between age and PLP concentration [38, 51, 58-61]; others, however, did not [32, 62, 63]. In a long-term study including 360 subjects ≥60 years of age, no decrease of the PLP concentration with advancing age was observed [52]. Possible causes discussed for lower PLP concentrations in elderly people comprise lower vitamin B₆ intake, reduced albumin synthesis, inflammation, and increased alkaline phosphatase activity [16, 36, 62]. However, age-related changes of vitamin B₆ bioavailability or metabolism have not been proven so far [36, 62]. In a study including 12 subjects aged >60 years, the vitamin B₆ intake level was estimated that was required at different protein intakes to regain the baseline plasma PLP concentration after a low-vitamin B₆ diet [50]. In 2 women and 1 man who had a daily protein intake of 0.8 g/kg body weight, this required a vitamin B₆ intake of 1.33 mg/day (women) and 1.36 mg/day (men), respectively. Due to the small sample size and the fact that baseline PLP concentrations exceeded 30 nmol/L [50], the RI for adults above 65 years of age is 1.4 mg/day for women and 1.6 mg/day for men (Table 1) and thus corresponds to the reference values of younger adults.

No data from balance studies are available regarding the vitamin B₆ requirement for children and adolescents. Therefore, the reference values for children and adolescents are based on the values compiled for adults and take into account differences in body weight, an allometric exponent, growth factors to consider the requirements for growth, and a coefficient of variation of 10% (addition of 20%) (Table 3). Growth factors at the different age groups were calculated from the protein requirement for growth in relation to the maintenance requirement according to the World Health Organization (WHO) [14, 64]. When using the age groups and reference body weights the D-A-CH reference values are based upon [1], the AR values range between 0.5 and 1.5 mg/day. The resulting RI values for vitamin B₆ range from 0.6 mg/day (for 1 to under 4-year-olds) to 1.4 mg/day for female adolescents and 1.6 mg/day for male adolescents (15 to under 19-year-olds) (Tables 1, 3). We want to point out that the calculated values for adolescents aged 15 to under 19 years are higher than those for adults (Table 3). However, as no evidence of a higher vitamin B₆ requirement for adolescents exists, the RI values for female and male adolescents aged 15 to under 19 years are assumed to be the same as for adults (Table 3).

The reference values for the intake of vitamin B₆ for infants aged 0 to under 4 months were derived based on the vitamin B₆ content of breast milk, which is considered to be the optimal diet for infants [65, 66]. The reference values for infants are therefore estimated values.

The average vitamin B₆ content from 2 studies of mature milk from lactating women not supplementing vitamin B₆ is 0.145 mg/L [67, 68]. Assuming an average breast milk intake of 750 mL/day [69], vitamin B₆ intake in an exclusively breastfed infant is approximately 0.109 mg vitamin B₆/day. This intake is sufficient to ensure a PLP concentration of ≥30 nmol/L [68]. Therefore, the estimated value for adequate vitamin B₆ intake is 0.1 mg/day for infants aged 0 to under 4 months.

For infants aged 4 to under 12 months, data on the average vitamin B₆ requirement are lacking. Thus, in line with the approach of the European Food Safety Authority (EFSA) [28], the vitamin B₆ reference value for infants of this age group is derived by averaging the results of the extrapolation from the estimated value for infants under 4 months of age and of the extrapolation from the AR for adults. Based on the estimated value for infants under 4 months of age (0.1 mg/day), considering differences in body weight [1] and the allometric exponent, the vitamin B₆ intake is 0.14 mg/day. However, based on the average vitamin B₆ requirement for adults (1.2 mg/day for women and 1.4 mg/day for men), considering differences in body weight, the allometric exponent, and the growth factor [1], the vitamin B₆ intake is 0.45 mg/day for female infants and 0.49 mg/day for male infants. By calculating the arithmetic mean, an estimated value for an adequate vitamin B₆ intake of 0.3 mg/day is derived (Table 4).

Plasma PLP concentration decreases during pregnancy; especially in the third trimester, concentrations <20 nmol/L were observed [70-72]. Possible reasons are haemodilution, enhanced PLP transfer to the foetus, increased alkaline phosphatase activity, and a higher requirement due to tissue growth [72, 73]. In line with the EFSA [28], the requirement during pregnancy is derived on the basis of body weight gain and average tissue content of vitamin B₆. Based on weight gain during pregnancy (16.7 g/day in the first, 60.6 g/day in the second, and 54.2 g/day in the third trimester [1]), a vitamin B₆ requirement of 15 nmol/g tissue growth (3.7 mg/kg) [30] and an average bioavailability of 75%, the AR is 1.3 mg/day in the first trimester and 1.5 mg/day in the second and third trimesters. After taking a coefficient of variation of 10% (addition of 20%) into account, the recommended vitamin B₆ intake for pregnant women is 1.5 mg/day in the first trimester and 1.8 mg/day in the second and third trimesters.

Vitamin B₆ concentration of breast milk is influenced by the mother’s diet [67]. Based on a vitamin B₆ secretion with breast milk of 0.1 mg/day (see the section Infants), an average bioavailability of 75% and an average breast milk intake of 750 mL/day of an exclusively breastfed infant, the AR in lactating women is 1.3 mg/day. Considering a coefficient of variation of 10% (addition of 20%), the recommended vitamin B₆ intake is 1.6 mg/day.

In the following, the available data on health-related aspects of vitamin B₆ are outlined without an evidence judgement. As most studies do not investigate vitamin B₆ alone but in combination with other B vitamins, it is sometimes not possible to evaluate the unique role of the different B vitamins.

Several epidemiological long-term studies repeatedly demonstrated inverse associations between vitamin B₆ status or vitamin B₆ intake and risk of depression [74], cardiovascular diseases [75], and cognitive impairment [76, 77]. Furthermore, an inverse association between vitamin B₆ status and inflammation has been described [41, 43, 78]. Meta-analyses of epidemiological studies indicate negative associations for both dietary vitamin B₆ intake and PLP concentration with the general risk of cancer and specifically the risk of cancers of the gastrointestinal tract such as the colorectum [79, 80], oesophagus, stomach, and pancreas [80]. Another meta-analysis indicates a negative association between the PLP concentration and the risk of breast cancer in postmenopausal women [81]. However, an association between vitamin B₆ intake and risk of breast cancer could not be proven [81].

So far, most intervention studies could not confirm preventive effects of vitamin B₆ regarding risk of cancer, depression, cardiovascular diseases, impairment of cognitive functions in older age or Alzheimer’s disease, or an effect on bone remodelling [80, 82-89]. Based on available data, a vitamin B₆ intake exceeding the reference values is not clearly associated with preventive effects on health in the general population.

Within the present revision of the D-A-CH reference values for the intake of vitamin B₆, the method of derivation has been changed: it is now based on balance studies using plasma PLP concentration as a biomarker to determine the AR. Reliable data on the vitamin B₆ intake levels that are sufficient to achieve the desirable plasma PLP concentration of 30 nmol/L are available for women, but not for children, adolescents, and men. Therefore, the reference values for children and adolescents were derived from the data for women considering differences in body weight and increased requirements in children and adolescents due to growth. Likewise, the reference values for men were derived based on the AR of women with consideration of differences in body weight. Due to the changed derivation, the reference values for some age groups are slightly higher than before. For pregnancy and lactation, the reference values are somewhat lower. However, higher intake levels for pregnant women compared to non-pregnant women are now recommended already in the first trimester, whereas previously higher intake levels were recommended from the beginning of the second trimester.

The derivation and magnitude of the revised reference values show similarities as well as differences to the derivation and magnitude of reference values set by the EFSA [28], Institute of Medicine (IOM) [90], and WHO [91]. For infants, the D-A-CH reference values for vitamin B₆ are in line with the reference values set by EFSA, IOM, and WHO, whereas for other age categories as well as for pregnant and lactating women, differences can be noted.

Similarities regarding the derivation include (1) the absence of connecting the vitamin B₆ requirements to the protein intake levels, (2) the use of the circulating PLP concentration as a biomarker for vitamin B₆ status, (3) the extrapolation of the ARs for children, adolescents, and pregnant/lactating women from the ARs of adults, (4) the vitamin B₆ content in human milk as a basis for deriving reference values for infants, (5) the higher AR for pregnancy and lactation due to foetus requirements and milk secretion, respectively, and (6) the nature of the reference values, that is, adequate intake and estimated values for infants and RIs for adolescents and adults.

The differences include, for instance, (1) cut-off values for defining adequate PLP concentrations, (2) age categories, reference body weights, and growth factors, (3) age at which sex-specific reference values were applied, (4) requirements of older adults, and (5) reference values depending on trimester. More precisely, IOM [90] and WHO [91] use a PLP concentration of ≥20 nmol/L to define an adequate vitamin B₆ status and also consider studies analyzing other vitamin B₆ biomarkers, such as activation coefficients of erythrocyte aspartate aminotransferase and erythrocyte alanine aminotransferase as well as homocysteine concentration and the renal excretion of tryptophan metabolites following a tryptophan loading test to derive ARs for adults, respectively. In contrast, EFSA bases the derivation of ARs on the achievement of a PLP concentration of ≥30 nmol/L [28] similar to the here presented derivation. With respect to the age groups, the D-A-CH reference values for children and adolescents are stratified into 7 age groups, whereas IOM [90] and WHO [91] use only 4 and EFSA [28] applies 5 age groups. This and the fact that different population studies were quoted as source of the reference body weights and growth factors explain why the age at which sex-specific reference values were set differs among the authorities. With reference to the above-mentioned study by Ribaya-Mercado et al. [50], IOM [90] and WHO [91] assume that older subjects require a higher vitamin B₆ intake to maintain a sufficient vitamin B₆ state compared to young adults and thus set age-specific reference values for adults. EFSA [28] abstains from setting age-specific reference values in adulthood but takes the results from Ribaya-Mercado et al. [50] into account by applying an AR of 1.3 mg/day for women and, after extrapolation, 1.5 mg/day for men. In contrast, D-A-CH nutrition societies conclude that the evidence is insufficient to set a higher AR and in consequence RI for older adults. In terms of pregnancy, the here presented D-A-CH reference values for pregnant women are the only ones that imply a trimester-specific vitamin B₆ intake, although EFSA [28], IOM [90], and WHO [91] acknowledge that the lower vitamin B₆ state is particularly evident in late gestation. In contrast to EFSA [28], IOM [90], and WHO [91], the D-A-CH reference values are based on trimester-specific weight gain during pregnancy taking into account that vitamin B₆ requirements rely on tissue growth.

According to representative data from the German National Nutrition Survey II, the median vitamin B₆ intake is 1.2 mg/day in women and 1.6 mg/day in men [92, 93]. This means that the recommended vitamin B₆ intake of 1.4 mg/day for women and 1.6 mg/day for men is not achieved by parts of the general adult population in Germany. Using the PRIs set by EFSA [28], that is, 1.6 mg/day for women and 1.7 mg/day for men, the situation appears even more critical, particularly for female subjects. However, irrespectively of the applied reference value, a vitamin B₆ intake below the RI does not necessarily represent a state of vitamin B₆ deficiency; rather it indicates a higher risk of suboptimal intake. The aim of reference values is the adequate nutrient supply of the healthy individuals in a population. The RI meets the requirements of ∼98% of the population. Thus, the RI exceeds the actual requirements of most people. Therefore, the RI is not suitable to distinguish between an adequate and inadequate nutrient intake on individual level or to diagnose a state of nutrient deficiency. For the examination of the probability that the usual intake of an individual is inadequate or to estimate the prevalence of inadequate intake in groups, the AR can be used [94, 95]. Referring to the AR, the above-mentioned median vitamin B₆ intake levels found in the German National Nutrition Survey II are in line with the AR for women and exceeded the AR for men. In light of this and the fact that the nutrient data base used to calculate the vitamin B₆ intake levels in the German National Nutrition Survey II considered predominantly non-fortified foods, vitamin B₆ is apparently no critical nutrient for the general population. However, the diagnosis of a deficiency should base on a comprehensive individual assessment including the measurement of recognized biomarkers, such as plasma PLP concentration for vitamin B₆.

Several protein-rich foods are good vitamin B₆ sources [16]. High levels of vitamin B₆ can be found in fish, such as sardines (cooked: 0.92 mg/100 g) and mackerels (cooked: 0.59 mg/100 g), as well as meat products, such as beef liver (cooked: 0.91 mg/100 g) [96]. In addition, cereals (e.g., whole-wheat flour: 0.46 mg/100 g), nuts (e.g., hazelnut: 0.66 mg/100 g and walnuts: 0.60 mg/100 g), dried fruits (banana: 0.93 mg/100 g and mango: 0.51 mg/100 g), and some vegetables (e.g., raw red pepper: 0.45 mg/100 g) contain relatively high amounts of vitamin B₆[96]. However, as pointed out previously, plant-based foods show sometimes high amounts of PNG [24]. For instance, high levels of PNG were noticed in frozen cauliflower/broccoli, cooked soy beans, raisins, fresh orange juice, and carrots with around two-thirds of the total vitamin B₆ content [24]. In contrast, the PNG proportion in walnuts, whole-wheat flour, bananas, and raw cauliflower is expected to be below 20% [24]. In a mixed diet, PNG usually accounts for approximately 15% of the total vitamin B₆ amount [26]. Thus, a mixed diet can provide adequate amounts of vitamin B₆ under the assumption of a bioavailability of 75%. The bioavailability was taken into account by the D-A-CH nutrition societies when setting the reference values. To achieve adequate vitamin B₆ supply, the daily intake of whole-grain products, vegetables, and fruits including a portion of nuts is recommended. Intake of fish once to twice a week and a low amount of meat can also contribute to cover the vitamin B₆ requirement.

The authors thank Professor Dr. Sabine Ellinger, Birte Peterson-Sperlich, and Prof. Dr. Bernhard Watzl for their valuable suggestions and contribution to the preparation of the revised reference values for vitamin B₆ intake.

The authors have no conflicts of interest to declare.

Dr. Alexandra Jungert received an honorarium from the German Nutrition Society (DGE) for developing the first draft of the dietary reference values for vitamin B₆ intake.

A.J. conducted the literature research and drafted the manuscript. J.L, K.-H.W., and M.R. revised the draft critically. All authors contributed to the conception of the manuscript and interpreted the data. All authors read and approved the final manuscript.