Vegans, and to a lesser extent vegetarians, have low average circulating concentrations of vitamin B12; however, the relation between factors such as age or time on these diets and vitamin B12 concentrations is not clear. The objectives of this study were to investigate differences in serum vitamin B12 and folate concentrations between omnivores, vegetarians and vegans and to ascertain whether vitamin B12 concentrations differed by age and time on the diet.
A cross-sectional analysis involving 689 men (226 omnivores, 231 vegetarians and 232 vegans) from the European Prospective Investigation into Cancer and Nutrition Oxford cohort.
Mean serum vitamin B12 was highest among omnivores (281, 95% CI: 270–292 pmol/l), intermediate among vegetarians (182, 95% CI: 175–189 pmol/l) and lowest among vegans (122, 95% CI: 117–127 pmol/l). In all, 52% of vegans, 7% of vegetarians and one omnivore were classified as vitamin B12 deficient (defined as serum vitamin B12 <118 pmol/l). There was no significant association between age or duration of adherence to a vegetarian or a vegan diet and serum vitamin B12. In contrast, folate concentrations were highest among vegans, intermediate among vegetarians and lowest among omnivores, but only two men (both omnivores) were categorized as folate deficient (defined as serum folate <6.3 nmol/l).
Vegans have lower vitamin B12 concentrations, but higher folate concentrations, than vegetarians and omnivores. Half of the vegans were categorized as vitamin B12 deficient and would be expected to have a higher risk of developing clinical symptoms related to vitamin B12 deficiency.
Although individuals who consume vegetarian or vegan diets may have a lower risk of cardiovascular disease (Key et al., 1999, 2009), there may also be a greater risk of developing nutritional deficiencies, because of the exclusion of meat and fish from the diet in vegetarians, and of all animal products in vegans. Vitamin B12 is naturally present only in foods of animal origin, and vegans who do not consume sufficient quantities of foods fortified with vitamin B12 such as some breakfast cereals, plant-based milks, soy products and yeast extract, or regularly take a vitamin B12 supplement, will have an increased risk of developing vitamin B12 deficiency.
Results from previous studies have shown that vegans have lower average serum concentrations of vitamin B12 in comparison with omnivores and vegetarians (Krajčovičová-Kudláčková et al., 2000; Herrmann et al., 2003; Majchrzak et al., 2006), with evidence from some (Hokin and Butler, 1999; Mann et al., 1999; Krajčovičová-Kudláčková et al., 2000; Hung et al., 2002; Herrmann et al., 2003), but not all (Haddad et al., 1999; Herrmann et al., 2001; Majchrzak et al., 2006), studies suggesting that serum vitamin B12 concentrations are also lower in vegetarians compared with omnivores. Several reports also indicate that a considerable proportion of vegans (Haddad et al., 1999; Waldmann et al., 2004) and vegetarians (Herrmann et al., 2003; Koebnick et al., 2004) have circulating concentrations of vitamin B12 indicative of depleted stores (serum vitamin B12 <150 pmol/l). Previous research in populations consuming a mixed diet suggests that, on an average, serum vitamin B12 concentrations are lower in older individuals (Baik and Russell, 1999); however, few studies have assessed this in older vegetarians and vegans. Moreover, it is not clear to what extent the length of adherence to a vegetarian or vegan diet is associated with serum concentrations of vitamin B12.
The objective of this study was to compare the serum vitamin B12 and folate concentrations and dietary intake of these micronutrients in British men consuming an omnivorous, vegetarian or vegan diet and to assess the associations of age and duration of adherence to a vegetarian or vegan diet with serum concentrations of vitamin B12.
The European Prospective Investigation into Cancer and Nutrition (EPIC)-Oxford cohort study recruited more than 65 000 participants between 1993 and 1999. Full details of the recruitment methods and study participants have been earlier described (Davey et al., 2003). The protocol for the EPIC-Oxford study was approved by the Royal College of General Practitioners’ Clinical Research Ethics Committee, the Central Oxford Research Ethics Committee and local research ethics committees, and all participants gave a written informed consent.
This study population sample consists of 689 men who provided a blood sample at recruitment to the EPIC-Oxford cohort between 1994 and 1997 and who had no history of cancer. To maximize the heterogeneity of dietary exposure, approximately equal numbers of men with different dietary habits were selected with no matching criteria: omnivores who reported consuming three or more servings of meat per week; lacto-ovo vegetarians who reported consuming soya milk and/or soya products; and vegans. There were 226 omnivores, 231 vegetarians and 232 vegans included in these analyses. The ethnicity of 96% of the study sample was White and the other 4% were Bangladeshi or Chinese.
Omnivores were defined as people reporting meat consumption. Vegetarians were defined as people who reported consuming dairy products (including milk, cheese, butter and yogurt) and eggs (including eggs in cakes and other baked foods) but no meat or fish. The majority of vegetarians were lacto-ovo vegetarians (88%) consuming both dairy products and eggs, with smaller proportions avoiding eggs (8%) or dairy products (4%). Vegans were defined on the basis that they did not consume any foods of animal origin (meat, fish, dairy products or eggs). The duration of adherence to a vegetarian or vegan diet was calculated as the age at recruitment minus the age at which the respondent last ate meat or fish, or any meat, fish, dairy product or eggs, respectively.
Assessment of dietary and lifestyle variables
All participants completed a validated semiquantitative 130-item food frequency questionnaire at baseline (Bingham et al., 1994, 1995). Regular use of a vitamin supplement over the past 12 months was assessed in the food frequency questionnaire and participants were categorized on the basis of whether they regularly used a vitamin B12 supplement or a folate supplement. The daily intakes of vitamin B12 and folate obtained from supplements were also calculated. The 12 men who did not provide any information on the use of vitamin B12 or folate supplements were categorized as nonusers for each of these variables.
Participants self-reported their height and weight, and Quetelet's body mass index (BMI; weight (kg)/height (m2)) was calculated. Participants were further characterized by their smoking status (‘never’, ‘former’ ‘current’) and level of education (‘some secondary school’; ‘higher secondary school’, ‘university degree or equivalent’).
After recruitment, blood was collected at local general practice surgeries into 10-ml Safety-Monovettes (Sarstedt, Nümbrecht, Germany). The samples were sent by post in sealed containers at ambient temperature to the EPIC laboratories in Norfolk, where they were centrifuged and aliquots were stored in liquid nitrogen (−196 °C) until analysis. Concentrations of vitamin B12 and folate in serum were measured with the use of a Quantaphase II B12/Folate Radioassay (Bio-Rad, CA, USA) at the Department of Clinical Biochemistry, King's College Hospital, London, UK. The interbatch coefficients of variation were 7.9% for vitamin B12 and 7.7% for folate. Serum concentrations of vitamin B12 less than 118 pmol/l and serum folate concentrations less than 6.3 nmol/l were used as cutoff points to categorize deficiency as used in the United Kingdom National Diet and Nutrition Survey (Ruston et al., 2004).
In 2001, the vegan men who had a serum vitamin B12 measurement were contacted and asked to provide another blood sample. Out of the 227 surviving vegan men contacted, another blood sample was collected from 65 (29%) of them. Full-blood counts were made using a Sysmex cell counter (Sysmex, Milton Keynes, UK); serum vitamin B12, serum holotranscobalamin (holoTC), plasma methylmalonic acid (MMA) and plasma total homocysteine (tHcy) concentrations were measured as described in detail (Lloyd-Wright et al., 2003).
All analyses were performed using Stata Statistical Software (release 9; Stata-Corp LP, TX, USA). Where necessary, dietary variables and serum vitamin B12 and folate concentrations were log transformed to approximate a normal distribution. To assess the effect of duration of adherence to a vegetarian or vegan diet on serum vitamin B12 concentrations, vegetarians and vegans were divided into four categories (5, 6–10, 11–15, 16 years) according to the time they had been on these diets. The geometric mean and 95% CI for serum B12 were calculated for nine diet-duration groups (omnivores, four groups of vegetarians and four groups of vegans) using multiple linear regression. Categories of BMI (<22.5, 22.5–24.9, 25 kg/m2), use of a vitamin B12 supplement, smoking status, level of education, number of days the blood sample was in the post (1, 2 or 3 days) and alcohol consumption (quartiles) were included as covariates, but only BMI, education and supplement use were included in the final model. Similar analyses were performed to investigate the association of age and education with serum vitamin B12 concentrations. P values for trend for the association between time of adherence to the diet and age with serum concentrations of vitamin B12 were assessed separately for each diet group by treating age and duration on the diet as continuous variables in the regression models. All tests were two-tailed and differences were regarded as statistically significant at P<0.05.
The mean age of participants was 47 years; omnivores were on an average 7 years older than vegetarians and 10 years older than vegans (Table 1). Vegans had the lowest mean BMI and a greater proportion of omnivores were categorized as having some secondary school level of education compared with vegetarians and vegans. The median time that vegetarian men had adhered to their diet was 11 years, and for vegan men it was 7 years. Among omnivores, 4% regularly took a supplement containing vitamin B12 compared with 19% of vegetarians and vegans.
The mean intake of dietary vitamin B12 among omnivores not taking a vitamin B12 supplement was 8.8 μg, which was almost five times greater than the mean intake in vegetarians (P<0.001) and 36 times higher than the mean intake in vegans (P<0.001, Table 2). Only 3% of vegan men not taking a vitamin B12 supplement reported a dietary intake of vitamin B12 above the UK reference nutrient intake of 1.5 μg/day compared with 31% of vegetarians and all omnivores. Among vitamin B12 supplement users, 89% of vegetarians and 63% of vegans met the reference nutrient intake for vitamin B12 intake. Vegans not taking a folic acid supplement had a significantly higher intake of folate than both vegetarians (P=0.001) and omnivores (P<0.001) and the mean intake of folate in vegetarians was higher than that in omnivores (P=0.014). Dietary intake of folate was above the reference nutrient intake (200 μg/day) in 96% of omnivores, 99% of vegetarians and 98% of vegans. All supplement users had an intake of folate (from diet and supplements) that was above the reference nutrient intake.
The mean serum vitamin B12 concentration in vegans was 33% lower than in vegetarians and 57% lower than in omnivores, and was 35% lower in vegetarians compared with omnivores (Table 3). In all, 52% of vegans and 7% of vegetarians had vitamin B12 concentrations below the cut-off point for biochemical deficiency (<118 pmol/l). A further 21, 17 and 1% of vegans, vegetarians and omnivores, respectively, had a serum vitamin B12 indicative of depletion (118–150 pmol/l). There was no significant difference in mean serum concentration of vitamin B12 between men who reported taking a vitamin B12 supplement compared with nonusers of supplements in any of the diet groups (results not shown).
The mean concentration of serum folate in vegans was 34% higher than in vegetarians and 88% higher than in omnivores (Table 3). Very few men (<1% of omnivores) were categorized with biochemical folate deficiency (<6.3 nmol/l). Vegetarians who reported taking a folate supplement had significantly higher mean serum folate concentration compared with nonsupplement users (P=0.015) but mean serum folate concentrations between supplement and nonsupplement users did not differ for omnivores (P=0.738) and vegans (P=0.072, results not shown).
Among participants not taking a supplement, dietary intake of vitamin B12 was significantly correlated with serum vitamin B12 concentrations (Spearman's ρ=0.72, P<0.001). The correlation between dietary folate intake and serum concentrations of folate was weaker but statistically significant (ρ=0.12, P=0.002).
The adjusted mean concentrations of serum vitamin B12 by time of adherence to a vegetarian or vegan diet and by age are shown in Figure 1. There was no significant association between the length of adherence to a vegetarian or vegan diet and serum vitamin B12 concentrations. There was also no significant association between age or education and serum concentrations of vitamin B12 in any diet group (results for education not shown). All results were similar when restricted to nonusers of vitamin B12 supplements (results not shown).
Of the 65 vegan men with repeated measures of serum vitamin B12 taken approximately 6 years later, 34% were considered as biochemically vitamin B12 deficient (serum vitamin B12 <118 pmol/l) and 8% were categorized as depleted (118–150 pmol/l) (Table 4). There was a strong positive association between serum vitamin B12 concentration and holotranscobalamin (holoTC) (r=0.77, P<0.0001) and an inverse association with methylmalonic acid and total homocysteine (r=−0.74 and r=−0.73, respectively; P<0.0001 for both). Among the 22 men who had serum vitamin B12 concentrations below 118 pmol/l, 82% had concentrations of holoTC that would indicate vitamin B12 deficiency (<35 pmol/l) but only 32% had a combination of a methylmalonic acid greater than 0.75 μmol/l and total homocysteine greater than 15 μmol/l that would identify men who were likely to have a vitamin B12 deficiency.
The results from this study show that mean serum concentrations of vitamin B12 reflected the level of intake of meat and animal products; omnivores had the highest concentrations of serum vitamin B12 and vegans the lowest. In addition, over half of the men consuming a vegan diet and 7% of men consuming a vegetarian diet were categorized—according to their serum vitamin B12 concentrations—as being vitamin B12 deficient. Very few men were categorized as being folate deficient (only two omnivores).
The finding that vegans have lower serum concentrations of vitamin B12 and a greater prevalence of biochemical vitamin B12 deficiency than both omnivores and vegetarians is in agreement with results from several other studies (Krajčovičová-Kudláčková et al., 2000; Obeid et al., 2002; Majchrzak et al., 2006). However, the results from some but not all studies have shown that vegetarians have lower vitamin B12 concentrations compared with omnivores and a greater proportion of vegetarians had serum vitamin B12 concentrations indicative of deficient or depleted levels (Haddad et al., 1999; Hokin and Butler, 1999; Mann et al., 1999; Krajčovičová-Kudláčková et al., 2000; Herrmann et al., 2001, 2003; Hung et al., 2002; Majchrzak et al., 2006). The results from this study do not support the idea that vegetarians, because they consume some animal products, are not at risk of developing a vitamin B12 deficiency (Herbert, 1994).
These results showed no evidence that serum vitamin B12 concentrations decreased with increasing duration on a vegetarian or vegan diet, which is in agreement with results from one other small study (Rauma et al., 1995) but not with another, in which an inverse correlation between serum vitamin B12 concentrations and time on a vegan diet (r=−0.18, P=0.047) was reported (Waldmann et al., 2004). Evidence from an intervention study suggests that when the intake of vitamin B12 is restricted, serum concentrations of vitamin B12 may decline much more rapidly than previously believed. In 13 lacto-ovo vegetarians, mean serum vitamin B12 concentrations were reduced from 345 to 226 pmol/l within 2 months of consuming a vegan diet (Crane et al., 1994). Only two men in this study had recently started consuming a vegan diet (<1 year) and therefore the effect on vitamin B12 concentrations of a short- term dietary change could not be assessed. It is possible that once the amount of vitamin B12 declines to a certain level, serum concentrations are maintained despite a very low intake over a prolonged period through a number of mechanisms that might include an increased absorption of vitamin B12 in the gut, reduced vitamin B12 excretion and an increased capacity to absorb recycled vitamin B12 from the bile through the enterohepatic circulation (Herbert, 1994).
Another group at risk of becoming vitamin B12 deficient is the elderly because of a greater prevalence of gastric atrophy and hypochlorhydria, both of which lead to an insufficient absorption of food-bound vitamin B12 (Baik and Russell, 1999). Indeed, others have reported a decrease in average serum vitamin B12 concentrations with advancing age in elderly populations (>65 years) (Nilsson-Ehle et al., 1991; Wahlin et al., 2002; Green et al., 2005; Loikas et al., 2007; Den Elzen et al., 2008). There was no evidence of an age-related decline in serum concentrations of vitamin B12 among any of the diet groups in this study, which is similar to that reported in the United Kingdom National Diet and Nutrition Survey (Ruston et al., 2004); this may be because these studies did not include many elderly participants.
The finding that over 95% of vegans and 31% of vegetarians who were not using supplements failed to meet the reference nutrient intake for daily vitamin B12 intake (1.5 μg/day from foods) was similar to that reported by others (Draper et al., 1993; Koebnick et al., 2004). Furthermore, even though vegetarian and vegan men who reported taking a vitamin B12 supplement had a higher intake of vitamin B12, serum concentrations of vitamin B12 were not affected. It is possible that supplement use was not accurately reported, some of the vitamin B12 supplements taken contained a type of inactive plant- based vitamin B12 (Watanabe, 2007) or that a proportion of men taking a vitamin B12 supplement had been recently diagnosed with a vitamin B12 deficiency. There seems to be a degree of awareness among these participants of the need to supplement vegan and vegetarian diets, in as much as 20% of vegans and vegetarians reported taking vitamin B12 supplements regularly. However, because there was little difference in serum vitamin B12 between supplement and nonsupplement users, it may be necessary to improve the understanding of the need to regularly consume supplements containing adequate amounts of the active form of vitamin B12.
Subclinical abnormalities are thought to emerge when serum concentrations of vitamin B12 fall below 111 pmol/l (Herbert, 1994). Over half of the vegans had serum vitamin B12 concentrations below 118 nmol/l; thus, these men have a higher risk of developing clinical symptoms related to vitamin B12 deficiency. Evidence of severe neurological damage as result of chronic vitamin B12 deficiency in vegans and vegetarians is rare; however, several case reports have shown neurological impairment that includes irreversible degeneration of the spinal cord, as well as difficulty in walking and handling utensils (Takahashi et al., 2006; Brocadello et al., 2007). There are very few studies that have looked for milder neurological symptoms among vegetarians and vegans who are vitamin B12 deficient.
In this study, serum concentration of vitamin B12 was used as a biomarker of vitamin B12 status, although recent evidence suggests that serum holoTC may be a more sensitive and specific biomarker of vitamin B12 status (Hvas and Nexo, 2005; Clarke et al., 2007). HoloTC is the fraction of vitamin B12 bound to transcobalamin that delivers vitamin B12 to cells that synthesize DNA and is thus considered the physiologically active component of vitamin B12 (Gibson, 2005). The concentrations of holoTC, methylmalonic acid and total homocysteine were measured in a subgroup of study participants; approximately 80% of men with serum vitamin B12 concentrations below the cutoff point of 118 pmol/l had values for holoTC that were indicative of vitamin B12 deficiency. On the other hand, only a third of these men had values for both methylmalonic acid and total homocysteine above the cut-off points considered as biochemically vitamin B12 deficient. Moreover, although the cut-off points used in this study to identify those at risk of vitamin B12 deficiency have been used elsewhere (Ruston et al., 2004), others have used a higher cutoff point (250 pmol/l) to define vitamin B12 deficiency (Lindenbaum et al., 1994). Therefore, it is possible that the true proportion of individuals with a vitamin B12 deficiency reported herein has been underestimated.
In conclusion, the results from this study show that vegetarians and vegans have much lower concentrations of serum vitamin B12 but higher concentrations of folate in comparison with omnivores. Mean serum vitamin B12 was not associated with the duration of adherence to a vegetarian or vegan diet, which may indicate that mechanisms that maintain circulating concentrations of vitamin B12 are upregulated in vegetarians and vegans. Further research into the health effects of vitamin B12 deficiency and depletion in vegans and vegetarians is warranted , and vegetarians and vegans should ensure a regular intake of sufficient vitamin B12 from fortified foods and/or supplements.
We thank all the participants in the EPIC-Oxford cohort for their invaluable contribution to the study, as well as the laboratory staff at King's College London, UK, for the measurement of serum vitamin B12 and folate concentrations. This study was conducted during tenure of a Girdlers’ New Zealand Health Research Council Fellowship (FLC) and we further acknowledge the Dutch Cancer Society (AMJG). The study was funded by Cancer Research UK; Medical Research Council, UK; and the European Commission.
About this article
European Journal of Clinical Nutrition (2014)