Fortification of cereal products with folic acid – the synthetic analogue to folate used in food fortification and nutritional supplements – was introduced in North America in 1998 to reduce the risk of pregnancies affected by neural tube defects. The mandatory fortification programme has caused a more than 100 % increase in median serum folate levels among the US population(Reference Pfeiffer, Johnson and Jain1). In European countries where no mandatory food fortification with folic acid or other B-vitamins has been introduced, low status of folate is believed to be common(Reference Dhonukshe-Rutten, de Vries and de Bree2). Besides the increased risk of neural tube defects, insufficient status of folate and/or related B-vitamins might contribute to the risk of several other diseases: CVD(Reference McCully3), neurological diseases(Reference Reynolds4) and certain cancers(Reference Sanjoaquin, Allen and Couto5–Reference Garcia-Closas, Castellsague and Bosch7).
The B-complex vitamin folate facilitates the transfer of one-carbon units and is essential as a cofactor or coenzyme in a variety of biological processes: for example, synthesis and repair of DNA, regulation of gene expression, amino acid metabolism, neurotransmitter synthesis, and the formation of myelin(Reference Lucock8). Vitamin B12 (cobalamin) is required as a cofactor for the two enzymes methionine synthase and methylmalonyl CoA mutase. Methionine synthase catalyses the remethylation of homocysteine to methionine which is also dependent on folate; thus, folate and vitamin B12 intersect in this metabolic process(Reference Banerjee9). Severe folate deficiency leads to megaloblastic anaemia while vitamin B12 deficiency may cause neurological symptoms in the form of myelin degeneration and irreversible cognitive impairment that is seen with or without the coincidence of anaemia.
The Danish legislation regarding food fortification has been very restrictive compared with most other European countries and no fortification with folate or vitamin B12 has been allowed until now. However, little is known about the actual prevalence of low folate and vitamin B12 status in the general Danish population, and it is not known if subgroups of the population owing to lifestyle factors or genetics are particularly at risk of insufficient folate and/or vitamin B12 status. Since there are no food items fortified with folate or vitamin B12 on the Danish market, Denmark offers a unique setting for studies of the effect of lifestyle and genetics on folate and vitamin B12 status. The aim of the present study was to evaluate the folate status and the vitamin B12 status of a general adult population in Denmark. In addition, we investigated the influence of different lifestyle factors and polymorphisms of genes involved in the metabolism of folate and/or vitamin B12 on serum levels of folate and vitamin B12.
Methods
Study population
The individuals included in the present study were participants in the Inter99 study. The study design and characteristics of the participants have been described in detail elsewhere(Reference Jørgensen, Borch-Johnsen and Thomsen10). In brief, an age- and sex-stratified random sample of 13 016 men and women born in 1939–40, 1944–5, 1949–50, 1954–5, 1959–60, 1964–5 and 1969–70 and living in eleven municipalities in the South-Western part of the former Copenhagen County was drawn from the Civil Registration System and invited to a health examination. A total of 12 934 individuals were eligible for invitation of whom 6784 (52·5 %) participated. In general the participation rate was higher in women than in men, and it increased with increasing age. In addition, non-responders had more hospital admissions related to chronic diseases such as diabetes and CVD(Reference Jørgensen, Borch-Johnsen and Thomsen10). The examinations of participants in the present study were completed from March 1999 to January 2001. The health examination included a self-administered questionnaire, a physical examination and various blood tests.
The Inter99 study has been approved by the ethics committee of Copenhagen County (KA 98 155) and the National Board of Health and the study was registered in the ClinicalTrials.gov (NCT00289237). An informed consent has been obtained from all participants.
Measurements of folate and vitamin B12
After an overnight fast blood samples for measurements of serum levels of folate and vitamin B12 were collected into tubes and left for clotting before centrifugation. Serum samples were stored at − 20°C until the analyses were performed in 2008. Serum folate and vitamin B12 concentrations were measured by using a competitive chemiluminescent enzyme immunoassay (Immulite® 2000 System; Siemens Medical Solutions Diagnostics, Los Angeles, CA, USA). Serum levels of folate and vitamin B12 were successfully measured for 6371 (93·9 %) and 6216 (91·6 %) of the participants, respectively.
We used the following standard cut-offs: < 6·8 nmol/l for low serum folate and < 148 pmol/l for low serum vitamin B12(Reference Pfeiffer, Caudill and Gunter11). In addition, we used a lower cut-off value for folate deficiency ( < 4·0 nmol/l) in some analyses. This cut-off value is used to define folate deficiency in some clinical settings in Denmark.
Genotyping
Participants were genotyped for the following genetic variants with potential relevance to the metabolism of folate and/or vitamin B12: two single nucleotide polymorphisms (SNP) in the methylenetetrahydrofolate reductase gene (MTHFR-C677T (rs1801133) and MTHFR-A1298C (rs1801131)), one in the methionine synthase gene (MTR-A2756G (rs1805087)), one in the methionine synthase reductase gene (MTRR-A66G (rs1801394)), one in the betaine:homocysteine methyltransferase gene (BHMT-G742A (rs3733890)) and two polymorphisms in the transcobalamin gene (TCN2-C6776G (rs1801198) and TCN2b (rs9606756)). The above-mentioned SNP were genotyped by TaqMan allelic discrimination (KBiosciences, Hoddesdon, Herts, UK). All genotyping success rates were above 96·2 % with a mismatch rate below 0·52 % in 384 duplicate samples and none of the observed genotype distributions deviated significantly from Hardy–Weinberg equilibrium (P>0·2).
Lifestyle factors
BMI was calculated as weight divided by height squared. Height and weight were measured wearing light clothes and no shoes. Participants were divided into categories based on their BMI according to the following criteria recommended by WHO(12): underweight (BMI < 18·5 kg/m2); normal range (BMI ≥ 18·5 to 25 kg/m2); overweight (BMI ≥ 25 to 30 kg/m2); obese (BMI ≥ 30 kg/m2). The self-administered questionnaire provided information on several lifestyle factors. Smoking status was defined in four categories: never smokers; ex-smokers; occasional smokers ( < 1 g tobacco or cigarettes per d); daily smokers ( ≥ 1 g/d). Total alcohol intake was calculated by summation of self-reported weekly intake of beer, strong beer, wine, dessert wine and spirits. Thus, one beer, one glass of wine, or one glass of spirits was approximated to one standard unit (defined as 12 g pure alcohol) and a strong beer was calculated as 1·5 standard units. Five categories were defined from the calculated alcohol intake: 0 units/week; 1–7 units/week; 8–14 units/week; 15–21 units/week; 22 or more units/week. In addition, beer and wine intakes were divided into four categories: 0 units/week; 1–3 units/week; 4–7 units/week; 8 or more units/week. Five categories were defined from self-reported daily intake of coffee: 0 cups/d; 1–3 cups/d; 4–6 cups/d; 7–9 cups/d; 10 or more cups/d. A variable of total physical activity based on information on commuting and leisure time physical activity were used to form four groups of distinct levels of physical activity: 0–2 h/week; 2–4 h/week; 4–7 h/week; 7–12 h/week(Reference von Huth, Borch-Johnsen and Jorgensen13). Dietary habits were measured using a validated dietary quality score developed from a forty-eight-item FFQ. The dietary score was developed as a crude index of overall quality of the dietary habits and a scoring system was used to divide the participants into three groups: healthy, average, and unhealthy dietary habits(Reference Toft, Kristoffersen and Lau14). Dietary intake of folate and vitamin B12 was estimated by calculations based on a 198-item FFQ(Reference Toft, Kristoffersen and Ladelund15). The study participants were divided into quartiles from their estimated intake of folate and vitamin B12.
Statistical analyses
The statistical program SAS (version 9.2; SAS Institute Inc., Cary, NC, USA) was used for all analyses. For categorical outcome variables (low folate status and low vitamin B12 status), χ2 tests were used to assess differences between groups. Multivariable logistic regression models were used to estimate OR for the independent effects of different lifestyle factors on low serum folate and vitamin B12 when taking into account possible confounding by other considered variables. Since the total alcohol variable includes both beer and wine intake, which may be reverse in their associations with folate and vitamin B12 status, beer and wine intake were included in the adjusted models whereas total alcohol intake was omitted except from models estimating the effect of total alcohol intake itself. In these models, none of the other alcohol-related variables was included. Potential effect modifications by genetic variants were evaluated by assessing the P values of the respective interaction terms (for example, MTHFR-C677T × lifestyle) in the regression models corrected for multiple testing (Bonferroni). A priori we decided only to test first-order interactions between SNP and each of the considered lifestyle factors whereas gene–gene and lifestyle–lifestyle interactions were not evaluated. P values of likelihood ratio tests were used to test for statistical significance in all logistic regression analyses.
One-way ANOVA was used to detect differences in continuous outcome variables (serum concentrations of folate and vitamin B12) between study groups. To achieve normal distribution of the continuous outcome variables, measurements of serum folate and vitamin B12 were log-transformed. Multiple linear regression models were used for adjustment for potential confounding. Results were computed as percentage differences compared with the reference group corresponding to the back-transformed β-coefficients from the linear regression analyses on log-transformed outcomes multiplied by 100(Reference Cole16). All P values are two-tailed and statistical significance was defined as P < 0·05.
Results
General characteristics of the study population are shown in Table 1. For serum folate, 95 % of the concentrations were in the range 3·4–26·3 (median 8·6) nmol/l, and 95 % of the serum vitamin B12 concentrations were between 128 and 702 (median 281) pmol/l. The prevalence of low serum folate and serum vitamin B12 concentrations stratified by sex and age group are shown in Table 2. The overall prevalence of low serum folate ( < 6·8 nmol/l) was 31·4 % and it was more common among men than women. In addition, low serum folate was inversely associated with age, with a prevalence of 43·4 % among the 30-year-olds and 18·3 % among those aged 60 years. The overall prevalence of folate deficiency ( < 4·0 nmol/l) was 5·1 %, and it was also inversely associated with age, while there was no significant difference between men and women. Correspondingly, the median and geometric mean values of the serum folate measurements were highest among those of oldest age and the measurements were significantly higher for women than men (data not shown). The inverse associations between low folate and folate deficiency and age were seen for both men and women when analyses were stratified by sex (data not shown).
Table 1 General characteristics of the population (n 6784)
Table 2 Serum folate and vitamin B12 according to age and sex in a general adult Danish population
* Prevalence in total sample of 6784 subjects.
† Differences in prevalence of low serum folate and vitamin B12 between groups were tested by χ2 statistical tests. P values relate to tests for differences between groups.
The overall prevalence of low serum vitamin B12 ( < 148 pmol/l) was 4·7 % and, in contrast to low serum folate, low serum vitamin B12 was significantly more common among women than men (8·4 v. 4·5 %) (Table 2). Low serum vitamin B12 was also inversely associated with age. However, this association was only seen for women, and the association disappeared when the statistical models were adjusted for folate status.
The influence of different lifestyle factors on the prevalence of low serum folate and folate deficiency is presented in Table 3. In crude logistic regression models low serum folate as well as folate deficiency were significantly associated with low folate intake, unhealthy diet, BMI (both underweight and obesity), low physical activity, high daily coffee intake, low total weekly alcohol intake, low weekly beer intake, low weekly wine intake, and daily smoking. In addition, the associations with folate intake, diet, physical activity, coffee intake and alcohol intake (total, beer, and wine) demonstrated dose-dependency. Regarding total alcohol intake, beer intake, and wine intake the dose-dependent associations were inverse, showing decreased risk of low serum folate and folate deficiency with increasing intake. When adjusting for confounding by sex, age and lifestyle factors in multiple logistic regression models, the effect of physical activity disappeared. However, the other associations between lifestyle factors and folate status remained significant even though the estimated OR to some degree were attenuated. Results from linear regression models with serum folate measurements as outcome (data not shown) confirmed the results from the logistic regression models in Table 3. In the crude models all the considered lifestyle factors were associated with folate status while the effect of physical activity disappeared when adjusting for confounding.
Table 3 Associations between lifestyle factors and serum folate in a general adult Danish population (n 6371)*
* P values relate to tests for differences between groups.
† Differences in prevalence of low serum folate between groups are tested by χ2 statistical tests.
‡ Adjusted logistic regression model including all considered lifestyle factors (except from alcohol intake) together with sex and age. To estimate OR for different categories of alcohol intake, beer and wine intake were excluded from the adjusted model. The effects of sex and age were also significant (P < 0·001) in the adjusted models. The adjusted models included 5368 participants with complete information on all considered variables.
§ Adjusted logistic regression model including all considered lifestyle factors (except from alcohol intake) together with sex and age. To estimate OR for different categories of alcohol intake, beer and wine intake were excluded from the adjusted model. The effects of sex and age were also significant (P = 0·012 and P = 0·019, respectively) in the adjusted models. The adjusted models included 5368 participants with complete information on all considered variables.
When looking at the results from unadjusted statistical models, significant associations were found between the prevalence of low serum vitamin B12 and high total alcohol intake, high beer intake, a daily coffee intake of 7–9 cups, and low estimated intake of vitamin B12 (Table 4). When adjusting for confounding by sex, age, and all considered lifestyle factors, only the effect from coffee intake and estimated intake of vitamin B12 remained significant. The effect of sex seen on serum vitamin B12 in Table 2 disappeared in the adjusted models. In addition, the highly significant associations with age indicated in Table 2 vanished when further adjusting for serum folate. There were significant associations between low serum folate and low serum vitamin B12 in the logistic regression model (P < 0·001) and between serum folate and serum vitamin B12 measurements in the linear regression model (P < 0·001).
Table 4 Associations between lifestyle factors and serum vitamin B12 in a general adult Danish population (n 6216)*
* P values relate to tests for differences between groups.
† Differences in prevalence of low serum vitamin B12 between groups were tested by χ2 statistical tests.
‡ Adjusted logistic regression model including all considered lifestyle factors (except from alcohol intake) together with sex and age. To estimate OR for different categories of alcohol intake, beer and wine intake were excluded from the adjusted model. The effect of age was also significant (P = 0·007) in the adjusted models while the sex effect was no longer significant (P = 0·179). When further adjusting for serum folate, the effect of age disappeared too (P = 0·106) while the other associations were not affected. The adjusted models included 5243 participants with complete information on all considered variables.
The only SNP showing significant association with serum folate and serum vitamin B12 was MTHFR-C677T (Table 5). The TT genotype of this variant was associated with increased prevalence of low serum folate (OR 2·24; 95 % CI 1·85, 2·70; P < 0·001)) as well as low serum vitamin B12 (OR 1·78; 95 % CI 1·25, 2·54; P = 0·003)) compared with the CC genotype. Correspondingly, it is seen from Table 5 that the geometric mean values of serum folate and serum vitamin B12 were significantly lowered by the TT genotype. None of the other tested SNP was consistently associated with serum folate or serum vitamin B12 although the GG genotype of the MTR-A2756G seemed to be associated with a decreased prevalence of low serum folate (OR 0·68; 95 % CI 0·50, 0·92; P = 0·017). However, no significant effect was seen on the geometric mean values. As expected from the principles of Mendelian randomisation(Reference Ebrahim and Davey17), there were no associations between genotypes and the considered lifestyle factors and therefore no statistical models including adjustments for confounding by lifestyle factors were applied. In addition, no significant interactions between SNP and lifestyle factors were found after correction for multiple testing.
Table 5 Influence of genetic polymorphisms on serum folate and vitamin B12 in a general adult Danish population*
MTHFR, methylenetetrahydrofolate reductase; MTR, methionine synthase; MTRR, methionine synthase reductase; BHMT, betaine:homocysteine methyltransferase; TCN, transcobalamin.
* P values relate to tests for differences between groups.
† Differences in prevalence of low serum folate and vitamin B12 between groups were tested by χ2 statistical tests.
‡ Differences in geometric means between groups were tested by one-way ANOVA.
Discussion
The present study demonstrates that low folate status is common among Danish adults. Almost a third of the participants had serum folate levels < 6·8 nmol/l and levels < 4·0 nmol/l were found in serum from 5·1 % of the population, whereas the overall prevalence of low serum vitamin B12 in the present study was 4·7 %. We found that serum folate was associated with nearly all the considered lifestyle factors in addition to the MTHFR-C677T polymorphism whereas serum vitamin B12 was only related to diet, BMI and coffee intake besides the MTHFR-C677T polymorphism.
The serum concentrations of folate measured in the present study may have been influenced by decomposition of folate during storage of the serum samples at − 20°C. In contrast, vitamin B12 is thought to be more stable and resistant to handling and storage(Reference Drammeh, Schleicher and Pfeiffer18, Reference Ocke, Schrijver and Obermann-de Boer19). From a number of studies, folate in serum samples is known to be unstable at room temperature wherefore folate measurements are vulnerable to sample processing and delayed freezing(Reference Drammeh, Schleicher and Pfeiffer18, Reference Hannisdal, Ueland and Eussen20). In addition, a few studies have demonstrated that folate also deteriorates over time at − 20°C(Reference Ocke, Schrijver and Obermann-de Boer19, Reference Lawrence, Umekubo and Chiu21) and thus the prevalence of low serum folate may be overestimated in the present study. In addition, serum folate levels show variation due to recent intake of folate, whereas erythrocyte folate is more stable and reflects average body folate status. Therefore erythrocyte folate would probably have been a better indicator of folate status in the present study, but such measurements were not available. In some clinical settings in Denmark, high total homocysteine (>15 μmol/l) is used to identify patients with likely folate deficiency. We have previously reported data on total homocysteine measurements in a randomly selected subgroup (n 2788) of the population included in the present study, and the prevalence of high total homocysteine in that subgroup was 4·6 %, which was very similar to the prevalence of serum folate concentrations < 4·0 nmol/l (4·8 %) (data not shown). In addition, there was a highly significant association between low serum folate and high total homocysteine (P < 0·0001). By assuming high total homocysteine (>15 μmol/l) as the ‘gold standard’ for defining inadequate folate status, the specificity and sensitivity of using low serum folate ( < 4·0 nmol/l) as an indicator of inadequate folate status were 97 and 37 %, respectively (OR 18·3; 95 % CI 12·0, 27·8).
As discussed below, the associations found in the present study may be influenced by the lack of information about use of supplements, which is a limitation of the present study. On the other hand, the reported associations between vitamin levels and lifestyle or genetic factors are probably not affected by the potential decomposition of folate during storage since the decline in absolute concentrations is thought to be non-differential and thus independent of genetic and lifestyle factors(Reference Ocke, Schrijver and Obermann-de Boer19). In addition, the previously reported data on total homocysteine showed associations with lifestyle factors and the MTHFR-C677T genotype resembling those found for serum folate in the present study(Reference Husemoen, Thomsen and Fenger22).
Lifestyle
As expected, serum folate measurements were positively associated with estimated dietary intakes of folate and consumption of a healthy diet(Reference Dhonukshe-Rutten, de Vries and de Bree2, Reference Green23). In accordance with other studies, we found an association between smoking and low serum folate(Reference Cafolla, Dragoni and Girelli24–Reference Vardavas, Linardakis and Hatzis26). The exact mechanisms behind the effect of smoking are not identified but smoking may inhibit enzymes such as methionine synthase(Reference de Bree, Verschuren and Kromhout27) or may interact with the remethylation of homocysteine to methionine and thereby possibly alter the ability of the cell to store and metabolise folate(Reference Piyathilake, Macaluso and Hine25).
The dose-dependent inverse association between serum folate and coffee intake is also consistent with another recent report(Reference Ulvik, Vollset and Hoff28), and in several other studies, high coffee intake has been associated with high total homocysteine indicating low folate status(Reference Husemoen, Thomsen and Fenger22, Reference de Bree, Verschuren and Kromhout27, Reference Rasmussen, Ovesen and Bulow29, Reference Nygard, Refsum and Ueland30). However, the present results indicate that the effect of coffee does not manifest itself at low-range folate concentrations since the prevalence of serum folate < 4 nmol/l was not affected by coffee consumption to the same degree. In fact, the results suggested a weak beneficial effect of a low to moderate coffee intake compared with no coffee intake. Our findings support a recent report by Ulvik et al. (Reference Ulvik, Vollset and Hoff28) who concluded that coffee consumption preferentially affects the upper, but not the lower, part of the B-vitamin concentration distributions(Reference Ulvik, Vollset and Hoff28). The mechanisms underlying the effect of coffee consumption are largely unknown. Caffeine has been proposed to inhibit the conversion of homocysteine to cysteine by acting as a vitamin B6 antagonist(Reference Grubben, Boers and Blom31). Another hypothesis is that coffee consumption simply causes low folate status by increasing the loss of folate by urinary excretion(Reference Ulvik, Vollset and Hoff28).
Studies examining relationships between alcohol consumption and serum folate or total homocysteine have been inconsistent and the overall results indicate that the relationship is complex(Reference de Bree, Verschuren and Kromhout27). In the present study, we found a dose-responsive positive association between serum folate and both total alcohol intake and beer intake. Consumption of wine was less strongly associated with serum folate. The present results are in line with associations between total homocysteine and alcohol intake previously reported in a subgroup of this cohort(Reference Husemoen, Thomsen and Fenger22). Consumption of alcohol has been suggested to be associated with total homocysteine in a J-shaped manner(Reference de Bree, Verschuren and Kromhout27), but in a recent randomised intervention study even a moderate alcohol (red wine or vodka) intake was associated with elevated total homocysteine and decreased levels of folate and vitamin B12(Reference Gibson, Woodside and Young32). In another intervention study, van der Gaag et al. (Reference van der Gaag, Ubbink and Sillanaukee33) reported decreased folate levels after intake of spirits but they found no effect on vitamin B12. After all, the effect of alcohol seems to depend on the type of alcoholic beverage consumed(Reference de Bree, Verschuren and Kromhout27, Reference van der Gaag, Ubbink and Sillanaukee33), and B-vitamins present in beer(Reference van der Gaag, Ubbink and Sillanaukee33) may to some degree be responsible for the positive effect of beer drinking on serum folate found in the present study. In addition, this may also explain the association with total alcohol since beer constitutes a major part of the total alcohol intake in this population.
Regarding vitamin B12, we found significant positive effects of estimated intake of the vitamin and a healthy diet. A correlation between vitamin B12 intake and vitamin B12 status has previously been shown even though the results are not consistent(Reference Dhonukshe-Rutten, de Vries and de Bree2). Besides, the only factors affecting serum vitamin B12 in the present study were obesity and coffee intake. In contrast to serum folate, vitamin B12 seemed to be positively associated with coffee consumption. These results do not support previous reports from observational(Reference Ulvik, Vollset and Hoff28) and randomised intervention(Reference Christensen, Mosdol and Retterstol34, Reference Verhoef, Pasman and Van Vliet35) studies showing no associations between coffee consumption and vitamin B12. However, an earlier study(Reference Desai, Zaveri and Antia36) indicated that coffee may increase the absorption of vitamin B12, which could explain our findings.
A major limitation of the present study is the lack of information on the use of dietary supplements. It has been estimated that about 50 % of the adult Danish population use multivitamin supplements normally containing both folate and vitamin B12, and 6 % use vitamin B products (either single supplements or complexes)(Reference Knudsen, Rasmussen and Haraldsdottir37). In a Danish study, supplement use was strongly associated with age and sex, being highest among elderly women(Reference Knudsen, Rasmussen and Haraldsdottir37), and, in general, supplement use seems to be associated with a healthier lifestyle profile and an already adequate nutritional intake(Reference Kirk, Cade and Barrett38). Users of dietary supplements have been found to be less likely to smoke, less likely to be obese, to drink less alcohol, and to exercise more than non-users(Reference Knudsen, Rasmussen and Haraldsdottir37, Reference Kirk, Cade and Barrett38). Therefore, unknown use of folate and vitamin B12-containing supplements may to some degree have influenced the associations with obesity (BMI ≥ 30 kg/m2), diet, age, sex and daily smoking found in the present study. However, the present results regarding lifestyle factors are consistent with previous findings in studies where only non-supplement users have been included or where adjustment for supplementary intake was feasible. Lack of information on use of supplements is unlikely to explain the somehow surprising effects of alcohol since individuals with a high alcohol intake are less likely to use dietary supplements than those with a lower intake so that potential confounding in this case rather would under- than overestimate the association.
It should also be acknowledged that the estimated prevalence of low folate and B12 status may be influenced by non-participation in the survey. For example, it has been indicated that non-responders were more likely to smoke and less likely to be overweight than those who participated in the examination(Reference Jørgensen, Borch-Johnsen and Thomsen10) and this may have influenced the estimates of associations in the present study.
Genetics
The MTHFR-C677T polymorphism has been extensively studied and the association between the TT genotype and low folate status is well documented(Reference Hustad, Midttun and Schneede39–Reference Al-Tahan, Sola and Ruiz42). Individuals with the TT genotype seem to be particularly susceptible to insufficient status of several B vitamins, and they may need to consume more folate to maintain serum folate levels similar to those found in individuals with the CC/CT genotypes(Reference Hustad, Midttun and Schneede39, Reference Nishio, Goto and Kondo40). Therefore, these individuals might be candidates for personalised nutritional recommendations. In contrast to previous studies(Reference Hustad, Midttun and Schneede39, Reference Al-Tahan, Sola and Ruiz42), we also found a significant association between the MTHFR-C677T polymorphism and vitamin B12.
The remaining SNP included in the present study have been considered as potential risk factors for neural tube defects due to their involvement in the metabolism of folate and some have previously been associated with altered folate status(Reference DeVos, Chanson and Liu41, Reference Gueant-Rodriguez, Rendeli and Namour43–Reference Gueant, Chabi and Gueant-Rodriguez45). In addition, mutations in the transcobalmin gene (TCN2) are known to alter the cellular availability of vitamin B12(Reference Gueant, Chabi and Gueant-Rodriguez45). However, we found no associations between these SNP and serum concentrations of folate and/or vitamin B12.
Conclusions and perspectives
In conclusion, we found that the prevalence of low serum folate was very common and was significantly associated with several common lifestyle and genetic factors in this general adult population where – by regulation – fortification of foods has not been allowed. Low serum vitamin B12 was less common and was only associated with a few lifestyle factors. Thus, the present results suggest that in populations without fortification many lifestyle and genetic factors influence folate levels and thereby that the vitamin status of the general population may be improved by introducing lifestyle changes. Thus, the findings may reinforce some current recommendations, for example, on healthy diet and smoking cessation. However, recommendations on supplementation to subgroups such as TT individuals of the MTHFR-C677T polymorphism are controversial and need further investigations in randomised studies.
Acknowledgements
We would like to thank the participants, all members of the Inter99 staff at the Research Centre for Prevention and Health, and The Steering Committee of the Inter99 study including Torben Jørgensen (principal investigator), Knut Borch-Johnsen (co-principal investigator), Hans Ibsen, Troels F. Thomsen, Charlotta Pisinger and Charlotte Glümer. The Inter99 study was supported by The Danish Medical Research Council, The Danish Centre for Evaluation and Health Technology Assessment, Novo Nordisk, Copenhagen County, The Danish Heart Foundation, The Danish Pharmaceutical Association, Augustinus Foundation, Ib Henriksens Foundation and Beckett Foundation. The present study was further supported by the Danish Agency for Science Technology and Innovation (grant no. 2101-06-0065) and Siemens Diagnostics, Denmark, who provided the Immulite Analyser for analyses of micronutrients.
B. H. T., L. L. N. H., A. L. and L. O. contributed to the development of the hypothesis and study design. T. J. was the principal investigator of the Inter99 study and responsible for data collection. M. F. performed the micronutrient analyses. B. H. T. performed the statistical analyses, wrote the first draft and coordinated the completion of the paper. All authors contributed to the interpretation of results, the revision of the manuscript, and have approved the final version of the paper.
The authors have no conflicts of interest to declare.
