|
|
||||||||
Hemostasis and Thrombosis |
1 Department of Medical and Surgical Critical Care, and3
Multidisciplinary Centre of Research on Food Sciences (G.R.A.), University of Florence, Florence, Italy.
2 Centro S. Maria agli Ulivi, Fondazione Don Carlo Gnocchi Onlus IRCCS-Impruneta, Florence, Italy.
4 Tuscany Regional Health Agency, Florence, Italy.
5 Geriatric Rehabilitation, Azienda Sanitaria di Firenze, Florence, Italy.
6 Longitudinal Studies Section, Gerontology Research Center, National Institute on Aging, National Institutes of Health, Baltimore, MD.
aAddress correspondence to this author at: Department of Medical and Surgical Critical Care, University of Florence, Viale Morgagni, 85, 50134 Florence, Italy. Fax 30-055-794-9418; e-mail annamaria.gori{at}unifi.it.
| Abstract |
|---|
|
|
|---|
Methods: Using data from 1320 participants in the population-based InCHIANTI study (586 men and 734 women; median age, 69 years; range, 21102 years) for whom complete data on folate, vitamin B6, inflammatory markers, 5,10-methylenetetrahydrofolate reductase (MTHFR) C677T sequence variant, and important covariates were available, we evaluated the association of inflammatory markers with circulating concentrations of vitamin B6 and folate, independently of dietary vitamin intake, circulating vitamin concentrations, and MTHFR C677T sequence variant.
Results: According to multiple linear regression analysis, C-reactive protein and interleukin-6 receptor were strongly and negatively associated with circulating vitamin B6 but not with folate concentrations, independent of age, sex, serum creatinine, serum albumin, total energy intake, smoking history, dietary nutrient intake, and circulating homocysteine and vitamin concentrations. Serum folate concentrations were related to MTHFR 677 TT genotype in persons with folate intake in the lowest tertile (<221.2 µg/day). Vitamin C and retinol intakes were strongly and positively associated with serum folate concentrations independent of age, sex, serum creatinine, serum albumin, total energy intake, smoking history, homocysteine plasma concentrations, dietary nutrient intakes, serum vitamin B6 and vitamin B12 concentrations, and MTHFR C677T sequence variant.
Conclusions: Low serum vitamin B6, but not serum folate, concentrations are independent correlates of the proinflammatory state, and both are influenced by antioxidant reserves.
| Introduction |
|---|
|
|
|---|
Evidence has accumulated on the crucial role of inflammation in the pathogenesis of atherosclerosis (8)(9). High C-reactive protein (CRP)1 concentrations are widely recognized as a strong independent risk factor for future cardiovascular events (10)(11)(12). Longitudinal studies have demonstrated that high interleukin-6 (IL-6) serum concentrations are associated with clinical progression of unstable angina and increased risk of myocardial infarction (13)(14)(15) and other cardiovascular events. IL-6 has greater predictive value than CRP as a cardiovascular risk factor (16).
Using data from the Framingham Heart Study, Friso et al. (17) demonstrated a strong relationship between impaired vitamin B6 status and a proinflammatory state, as indicated by CRP concentrations independent of homocysteine (Hcy) concentrations. These findings suggested that vitamin B6 may be implicated in some antiinflammatory mechanism, perhaps acting as a cofactor. In animal models, activated macrophages overexpress high-affinity folate receptors (18), suggesting that during an inflammatory state vitamin B6 may be mobilized from the liver and peripheral tissues to inflammation sites (19) and raising the interesting hypothesis that a prolonged proinflammatory state may lead to vitamin depletion, which perhaps contributes to a sustained chronic inflammatory response. Thus, relatively low concentrations of folate and vitamin B6 in older persons may be caused by the chronic mild proinflammatory state that characteristically affects this age group.
Using data from a large representative sample enrolled in the InCHIANTI study, we tested the hypothesis that the inflammatory state is associated with low vitamin B6 and folate concentrations independent of dietary intake and other potential confounders, including 5,10-methylenetetrahydrofolate reductase (MTHFR)2 C677T sequence variant.
| Materials and Methods |
|---|
|
|
|---|
All participants gave informed consent, and the study was approved by the Ethics Committee of the Italian National Institute of Research and Care of Aging. Blood samples were collected in the morning after the participants had been fasting for at least 8 h and sitting for at least 15 min. The analysis reported here is based on 1320 participants for whom complete data on folate, vitamin B6, inflammatory markers, MTHFR C677T sequence variant, and important covariates were available.
assays
We used commercial enzymatic tests (Roche Diagnostics, GmbH) to measure serum total cholesterol, HDL-cholesterol, and triglyceride concentrations. Serum LDL-cholesterol concentrations were calculated by the Friedewald formula. The interassay CVs were <3.8% for total cholesterol, <5% for HDL-cholesterol, and <2.5% for triglycerides. Lipid values were expressed according to the risk categories suggested by the National Heart, Lung, and Blood Institute Expert Panel on Detection, Evaluation. and Treatment of High Blood Cholesterol in Adults (21).
Folate and vitamins B6 and B12 were measured in serum samples obtained from blood collected in evacuated tubes without anticoagulant, centrifuged at 2000g for 10 min, and stored at 80 °C.
We measured vitamin B6 (pyridoxal-5-phosphate) with a commercial HPLC assay with fluorescence detection (Immun Diagnostik) and serum folate and vitamin B12 with a radioligand-binding assay (SimulTrac-SNB Radioassay; ICN Pharmaceuticals). The minimum detectable concentrations were 0.2 µg/L for vitamin B6, 0.6 µg/L for folate, and 75 ng/L for vitamin B12. The intraassay CVs were 2.8% for vitamin B6, 4.1% for folate, and 11% for vitamin B12, and the interassay CVs were 4.1% for vitamin B6, 7.1% for folate, and 12% for vitamin B12.
We measured circulating concentrations of inflammatory markers with commercial assays as described previously (22). Plasma Hcy concentrations were measured by a fluorometric polarized immunoassay method (IMX; Abbott Laboratories).
MTHFR C677T sequence variant analysis
We extracted genomic DNA from peripheral blood samples with a commercial reagent set (FlexiGene DNA reagent set; Qiagen GmbH). Screening for C677T MTHFR sequence variants was carried out on genomic DNA by microelectronic array technology (Nanochip Molecular Biology Workstation; Nanogen) (23). Briefly, genomic DNA was amplified with forward and reverse PCR primers (forward, 5'-biotin-TGA AGG AGA AGG TGT CTG CGG GA-3'; reverse, 5'-AGG ACG GTG CGG TGA GAG TG-3'). In addition, we designed 2 reporter oligonucleotidesone labeled with Cy3, specific for the wild-type nucleotide, and the other, labeled with Cy5, specific for the mutant nucleotide T (wild-type reporter, 5'-Cy3-TGA TGA AAT CGG-3'; mutant reporter 5'-Cy5-ATG ATG AAA TCGA-3')and 1 stabilizer oligonucleotide (5'-CTC CCG CAG ACA CCT TCT CCT TCA-3'). PCR products were desalted with the NucleoFast (96-well PCR plates; Macherey-Nagel) and electronically addressed to specific pads on the chip (NanoChip cartridge). After the stabilizer and reporter oligonucleotides were hybridized to the chip, the cartridge was placed in the reader, and the temperature was increased to 36 °C for discrimination. A fluorescence scan was performed, and the software directly assigned a genotype to each sample.
covariates
Average daily intakes of energy (kcal), alcohol (<30 vs >30 g/day), and vitamins were estimated by administration of the European Prospective Investigation into Cancer and Nutrition (EPIC) food frequency questionnaire. The EPIC food frequency questionnaire provides a detailed assessment of food consumption during the previous year through a large number of structured and precoded questions. Originally, the questionnaire was designed to be self-administered. However, in a pilot study we discovered that this method of administration provides ambiguous results for older adults because they are likely to misunderstand the questions. Thus, in the InCHIANTI study, the EPIC questionnaire was administered by the interviewers. The information provided by the questionnaire was transformed to average daily intake of macro- and micronutrients, including vitamins, by custom software that uses the table of food composition for epidemiologic study in Italy, edited by the European Institute of Oncology in 1998 (24)(25), for reference.
Smoking status was assessed by self-reported data. Pack-years, a measure of smoking exposure that combines intensity and duration, was calculated as (packs smoked per day) x (years of smoking).
statistical analysis
We performed all analyses with the SPSS statistical package, Ver. 11.5. P values <0.05 were considered statistically significant.
Log-transformed values for serum vitamin B6, folate, serum inflammatory markers, nutrient intake, and the square root of alcohol intake were used in the analyses and back-transformed for data presentation.
We calculated the MTHFR C677T genotype distribution across serum folate tertiles (
4.99, 5.08.16, and >8.16 nmol/L). Differences among serum folate tertiles were tested by general linear models adjusted for sex, age, serum creatinine, serum albumin, total energy intake, and smoking history.
Adjusted mean folate values according to MTHFR C677T genotype were estimated from a general linear model adjusted for age, sex, serum creatinine, serum albumin, smoking history, and total energy intake. We used multiple linear regression models to test the independent associations of inflammatory markers, nutrient intake, and other vitamins with vitamin B6 and folate. The model predicting folate also included the MTHFR genotypes as potential predictors. Variables found not to be associated with the outcome were removed from the final, most rigorous, regression model through a backward selection algorithm.
We studied the role of Hcy in the relationship between inflammation and vitamin B6 and serum folate concentrations by adding Hcy as a covariate in the final, most rigorous model predicting serum vitamin B6 and folate. The b coefficient estimated in the linear regression analysis indicates the expected mean change in logarithmically transformed vitamin B6 or folate associated with 1-unit changes in the independent variables. Therefore, for independent variables that are logarithmically transformed, a 1% change corresponds to a b% change in the dependent variable.
| Results |
|---|
|
|
|---|
|
correlates of serum vitamin b6 concentrations
In the multiple linear regression analysis, CRP and IL-6 receptor (IL-6r) were strongly and negatively associated with circulating vitamin B6 concentrations independent of age, sex, serum creatinine, serum albumin, total energy intake, smoking habits, dietary nutrient intakes, and circulating serum vitamin concentrations (Table 2
, model 1). After the inclusion of Hcy in the model (model 2), CRP, and soluble IL-6r (sIL-6r) remained significantly associated with vitamin B6 concentrations as well as age, alcohol and ß-carotene intakes, and serum folate, vitamin B12. and albumin concentrations (Table 2
, model 2).
|
correlates of serum folate concentrations
When we adjusted the model for age, sex, serum creatinine, serum albumin, total energy intake, and smoking habits, the prevalence of the MTHFR TT genotype was significantly (P <0.001) higher in the lowest than in the highest tertile of circulating serum folate (Table 3
). Serum folate concentrations were lower in participants carrying the MTHFR 677 TT genotype [median (5th95th percentiles), 5.71 (5.356.10) nmol/L] than in participants with the CT [6.59 (6.326.87) nmol/L] or CC [6.95 (6.527.41) nmol/L] genotype. The percentage of participants with low serum folate concentrations (<5 nmol/L) was significantly higher among those carrying the TT genotype (42.9%) than among those carrying the CC or CT genotype (28.9% and 31.7%, respectively; P <0.01).
|
Serum folate concentrations were related to MTHFR C677T sequence variant in participants with folate intake in the lowest tertile (<221.2 µg/day), but not in participants with folate intake in the highest tertiles (221.2292.2 and >292.2 µg/day). In particular, in participants in the lowest tertile of folate intake and carrying the TT genotype, serum folate concentrations were significantly lower than those found in participants with the CT or CC genotype (Fig. 1
).
|
In the multiple linear regression analysis, after adjustment for several confounders (age, sex, serum creatinine, serum albumin, total energy intake, smoking habits, dietary nutrient intakes, and circulating vitamin concentrations), vitamin C and retinol intakes and circulating serum concentrations of
-tocopherol were significantly and positively associated with circulating serum folate concentrations (Table 4
, model 1). Smoking history as well as circulating
-tocopherol concentrations and homozygosity for the MTHFR C677T sequence variant (TT genotype vs CT + CC genotypes) were significantly and inversely associated with circulating folate concentrations (Table 4
, model 1). Serum concentrations of CRP, IL-6, IL-6r, IL-1 receptor antagonist, and albumin and the leukocyte count were not significantly associated with serum folate concentrations [b (SE) log(CRP), 0.028 (0.035); log(IL-6), 0.011 (0.022); log(IL-6r), 0.019 (0.030); log(IL-1 receptor agonist), 0.019 (0.030); albumin, 0.034 (0.024); log(leukocyte count), 0.087 (0.069)]. After the inclusion of plasma Hcy concentrations in the model (model 2), vitamin C and retinol intakes, circulating serum concentrations of vitamin B6 and B12, sex, and smoking habits remained significant predictors of serum folate concentrations (Table 4
, model 2).
|
| Discussion |
|---|
|
|
|---|
Our findings support the notion already suggested by the Framingham study of a relationship between vitamin B6 and inflammation, and provide an original contribution to this literature by showing that such a relationship is not limited to CRP but also involves the sIL-6r. In addition, our data are in accordance with recent evidence of a link between inflammation and vitamin B6 not only in patients affected by inflammatory diseases such as rheumatoid arthritis and inflammatory bowel disease (19)(26), but also in stroke patients and in controls (27). Moreover, our findings are at variance with those observed in a subset of the ARIC study, in which leukocyte count but not CRP was a correlate of serum vitamin B6(28). Changes in sIL-6r concentrations have been determined in numerous clinical disorders (29), but the biological significance of sIL-6r within these pathologies remains unclear.
The mechanism by which inflammation is associated with low vitamin B6 concentrations has not been completely elucidated. It has been demonstrated that during a chronic inflammatory state, circulating and hepatic concentrations of vitamin B6 tend to decrease (19). Because vitamin B6 is involved in the repair and synthesis of nucleic acid and proteins, low vitamin B6 concentrations may reflect the increased consumption of this coenzyme in the accelerated synthesis of cytokines and antibodies and in accelerated cell proliferation.
Interestingly, we found that low vitamin B6 concentrations were independently associated with low ß-carotene intake and
-tocopherol concentrations, suggesting that an imbalance in the oxidative stress and antioxidant mechanisms may contribute to the decreases in vitamin B6 concentrations in inflammatory states.
We previously demonstrated that inflammation is strongly related to Hcy (22), and it is well known that vitamin B6 affects the serum concentration of Hcy. Thus, it was reasonable to hypothesize that the relationship between vitamin B6 and inflammatory state could be mediated by Hcy. However, the inclusion of Hcy in the linear regression model did not substantially attenuate the relationship between inflammatory markers and vitamin B6.
In our study, vitamin C intake and
-tocopherol concentrations were strongly associated with serum folate concentrations, independent of age, sex, smoking history, and folate intake, suggesting that noncompensated oxidative stress may contribute to the decrease in folate as well as serum vitamin B6 concentrations. The antioxidant molecule vitamin C is known to influence folate synthesis by enhancing formation of the polyglutamate derivatives from tetrahydrofolic acid and enhancing the transformation of folic acid into tetrahydrofolic acid (30). Interestingly, in Italian blood donors who were smokers, supplementation with folate plus vitamin C led to greater increases in serum and erythrocyte folate than did supplementation with folate alone (31).
By showing that MTHFR C677T sequence variant is an independent predictor of serum folate in older persons, especially those with low folate intake, our study confirms similar observations obtained in middle-aged European and North American populations (32)(33)(34).
In addition to the well-known influence on Hcy concentrations, folate may influence genomic DNA methylation reactions, particularly in participants with low serum folate concentrations and persons homozygous for the MTHFR C677T TT sequence variant (33).
The finding of the present investigation, demonstrating that low vitamin B6 is an independent correlate of proinflammatory state, may appear not to be consistent with the results of recent clinical trials showing no effect of folic acid, vitamin B6, and vitamin B12 supplementation in reducing cardiovascular events (35)(36). Nevertheless, all of the interventional trials confirmed the role of Hcy as marker of an increased risk of vascular events.
A limitation of this study is represented by the nature of our study (cross-sectional study), which did not allow us to determine the causative role of inflammation in inducing low circulating vitamin B6 concentrations.
In conclusion, we demonstrated that CRP and sIL-6r, 2 important biomarkers of inflammation, are independent correlates of vitamin B6 concentrations. Antioxidant vitamins, but not inflammatory markers, were independent correlates of serum folate. The authors of future studies looking at vitamins as modifiable risk factors for health outcomes must consider inflammation related to vitamin B6 and oxidative stress related to vitamin B6 and folate as potential confounders in the analyses.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Human gene: MTHFR, 5,10-methylenetetrahydrofolate reductase (NADPH). ![]()
| References |
|---|
|
|
|---|
T mutation on the relations among folate intake and plasma folate and homocysteine concentrations in a general population sample. Am J Clin Nutr 2003;77:687-693.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |