The purpose of this study was to investigate whether vitamin B6 supplementation had a beneficial effect on inflammatory and immune responses in patients with rheumatoid arthritis (RA).
This was a single-blind co-intervention study performed at the Division of Allergy, Immunology and Rheumatology of Chung Shan Medical University Hospital, Taiwan. Patients were diagnosed with RA according to the 1991 American College of Rheumatology criteria for RA. Patients were randomly allocated into two groups: control (5 mg/day folic acid only; n=15) or vitamin B6 (5 mg/day folic acid plus 100 mg/day vitamin B6; n=20) for 12 weeks. Plasma pyridoxal 5′-phosphate (PLP), serum folate, inflammatory parameters (that is, high-sensitivity C-reactive protein (hs-CRP), erythrocyte sedimentation rate (ESR), interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α)) and immune parameters (that is, white blood cell, total lymphocyte, T-cell (CD3), B-cell (CD19), T-helper cell (CD4), T-suppressor (CD8)) were measured on day 1 (week 0) and after 12 weeks (week 12) of the intervention.
In the group receiving vitamin B6, plasma IL-6 and TNF-α levels significantly decreased at week 12. There were no significant changes with respect to immune responses in both groups except for the percentage of total lymphocytes in the vitamin B6 group when compared with week 0 and week 12. Plasma IL-6 level remained significantly inversely related to plasma PLP after adjusting for confounders (β=−0.01, P=0.01).
A large dose of vitamin B6 supplementation (100 mg/day) suppressed pro-inflammatory cytokines (that is, IL-6 and TNF-α) in patients with RA.
Previous studies (Roubenoff et al., 1995; Galloway et al., 2000; Friso et al., 2001; Chiang et al., 2003; Talwar et al., 2003; Huang et al., 2005; Cheng et al., 2008) have indicated that lower plasma plasma pyridoxal 5′-phosphate (PLP) is associated with increased levels of pro-inflammatory cytokine (that is, tumor necrosis factor-α (TNF-α)) and inflammatory markers (that is, C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR)). Low vitamin B6 status due to inflammation might be caused by plasma PLP acting as a co-enzyme for the production of cytokines and other polypeptide mediators during the inflammatory response (Friso et al., 2001).
Rheumatoid arthritis (RA) is a chronic, autoimmune and systemic inflammatory disease that causes progressive joint destruction and is associated with a higher mortality rate compared with the general population (Ward, 2001; Gabriel et al., 2003). Studies have shown that patients with RA had lower circulating levels of vitamin B6 (that is, PLP) compared with healthy controls (Roubenoff et al., 1995, 1997; Chiang et al., 2005a; Woolf and Manore, 2008). Patients with RA were under chronic inflammatory conditions, which may possibly increase the use and metabolic turnover of plasma PLP. It was hypothesized therefore that supplementation of vitamin B6 would suppress inflammatory responses in patients with RA, but pro-inflammatory cytokine (TNF-α, interleukin-6 (IL-6)) production was not affected (Chiang et al., 2005b). The non-effect result might have been due to the lower dose (50 mg/day) and short treatment period (30 days) used in Chiang et al.'s (2005b) study. We thus postulated that a high dose of vitamin B6 supplementation (100 mg/day) and longer treatment period (12 weeks) would likely decrease the inflammatory responses of patients with RA.
Although the mechanism of plasma PLP's effect on immune responses has not yet been ascertained, plasma PLP was shown to be a significant indicator of immune responses in our critically ill patients (Huang et al, 2005). Supplementation of vitamin B6 seemed to be beneficial for improving immune responses (Casciato et al., 1984; Talbott et al., 1987; Folkers et al., 1993; Cheng et al., 2006). In addition to the relationship between vitamin B6 status and inflammatory responses, it would thus be useful to know whether a high dose of vitamin B6 supplement would also increase the immune response of patients with RA. The purpose of this study was to investigate whether vitamin B6 supplementation had a beneficial effect on inflammatory and immune responses in patients with RA. We, therefore, hypothesized that treatment with 100 mg pyridoxine for 12 weeks could change the responses of inflammation or immune function in patients with RA.
Materials and methods
This single-blind co-intervention study was conducted at the Chung Shan Medical University Hospital, which is a teaching hospital in central Taiwan. Adult patients had to fulfill the American College of Rheumatology criteria for RA (Hochberg et al., 1992). Clinical measurements, including patients’ visual analog scale, swollen and painful joint counts, disease activity score 28, rheumatoid factor and disease duration, were assessed by a rheumatologist. Patients were excluded if they were pregnant, had anemia (hemoglobin10 mg per 100 ml), thrombocytopenia (platelet count100 000 mm3), abnormal liver function (serum aspartate aminotransferase or alanine aminotransferase>40 IU/l), renal insufficiency (serum creatinine>1.4 mg per 100 ml), diabetes or cancer. Patients taking vitamins or other nutritional supplements were asked to stop for at least 1 month before entering the study. The study was approved by the Institutional Review Board of Chung Shan Medical University Hospital. Written informed consent was obtained from all subjects.
Patients’ demographic and health characteristics, family history and medication use were recorded. The body mass index (kg/m2) was calculated from height and weight measurements. Blood pressure (systolic and diastolic blood pressure (SBP and DBP)) was measured after a resting period of at least 5 min. Most of our RA patients (74%) were treated with methotrexate, currently the standard traditional disease-modifying anti-rheumatic drug, which is a folate antagonist. As folic acid supplements (5 mg) have been prescribed routinely to all patients receiving methotrexate for the treatment of RA (Whittle and Hughes, 2004), all patients were simultaneously supplemented with 5 mg/day folic acid to compensate for the folate-depleting side effect of methotrexate. The intervention dose (100 mg/day) of vitamin B6 is 62.5 times higher than the current Taiwan dietary reference intake (Department of Health, Taiwan, 2003) of 1.6 mg/day for men and women aged >51 years. Patients were then randomly allocated into two groups: control (5 mg/day folic acid only; n=15) or vitamin B6 (5 mg/day folic acid plus 100 mg/day vitamin B6; n=20). Intervention was administered for 12 weeks. The vitamin tablets were commercially available preparations (China Chemical & Pharmaceutical, Taipei, Taiwan and Johnson Chemical Pharmaceutical Works, Taipei, Taiwan). Patients were instructed to take tablets daily before breakfast and to refrain from using any other vitamin supplements during the study period. To ensure the compliance, investigators called the patients every week to remind them to take the tablets. Patients returned to the clinic to get additional supplements, at which time any unused tablets from the earlier 4 weeks were returned and counted. Subjects would be excluded if their compliance rate was under 80%. Patients were taught to maintain their usual diets and activities during the study period. To ensure that patients were keeping their usual dietary intake, they completed 24-h diet recalls when they returned to the clinic before the intervention (week 0) and at week 12. Nutritional composition was calculated using Nutritionist Professional software (E-Kitchen Business Corporation, Taiwan, 2002), and the nutrient database was based on the Taiwan food composition table.
Fasting blood samples were drawn from the patients by venipuncture on admission on the morning of day 1 (week 0) of the intervention and after 12 weeks (week 12). Serum or plasma was separated within 30 min after blood was collected and then frozen (−80 °C) until analysis. Serum white blood cell, total lymphocyte count and neutrophils were measured. Plasma PLP was determined by using high-performance liquid chromatography (HPLC) as previously described (Bates et al., 1999). The intra- and inter-assays of plasma PLP coefficient of variation were 1.1% (n=4) and 1.5% (n=6), respectively. Serum folate was analyzed by using standard competitive immunochemiluminometric methods on a Chiron Diagnostics ACS: 180 Automated Chemiluminescence System (Chiron Diagnostics Corporation, East Walpole, MA, USA). Automated high-sensitivity CRP (hs-CRP) measurement was performed using particle-enhanced immunonephelometry with an image analyzer (Dominici et al., 2004). The cut-off value of hs-CRP was set at 0.3 mg per 100 ml. The ESR was measured by using the Westergren method with automatic Sediplus S100 (Sarstedt, Numbrecht, Germany). The normal range of ESR was set at 0–9 mm/h for men and 1–20 mm/h for women. Interleukin-6 and TNF-α gene concentrations were assayed with the commercially available quantitative enzyme immunoassay kits (Bioscience, Camarillo, CA, USA). Lymphocyte subsets were analyzed by flow cytometry (FACS Calibur, San Jose, CA, USA) using fluorescent-labeled antibodies specific to the cell markers. All tests were performed within 48 h of sampling. The percentage of lymphocytes as a function of the total leukocyte count was determined by differential grating after two-color direct immunofluorescence staining. The T-lymphocyte (CD3) and B-lymphocyte (CD19) percentages were determined with fluorescein isothiocyanate-labeled CD3 (Leu4), clone SK7 and phycoerythrin-labeled CD19 (Leu-12), clone 4G7 (Becton Dickinson, Immunocytometry Systems, San Diego, CA, USA). T-helper (CD4) and T-suppressor (CD8) lymphocyte percentages were determined with fluorescein isothiocyanate–labeled CD4 (Leu 3a), clone SK3, and phycoerythrin-labeled CD8 (Leu-2a), clone SK1 (Becton Dickinson, Immunocytometry Systems). The following values were considered as the reference ranges in our laboratory: white blood cells, 5000–10 000/mm3; neutrophil, 3000–7000/mm3 or 60–70%; total leucocyte count, 1000–4000/mm3 or 20–40%; T-lymphocyte, 61–76%; B-lymphocyte (CD19), 6–22%; T-helper cell, 27–43%; and T-suppressor cell (CD8), 24–35% (Huppert et al, 1998).
Data were analyzed using the SigmaStat statistical software (version 2.03; Jandel Scientific, San Rafael, CA, USA). Differences in subjects’ characteristics and hematological measurements were analyzed by Student’s t-test or Mann–Whitney rank sum test between the two groups. For categorical response variables, differences between groups were assessed by χ2-test or Fisher's exact test. The paired t-test or Wilcoxon signed rank test was used to compare significant differences in demographic and health characteristics and the data for biochemical measurements within each group between week 0 and week 12. Multiple linear regression was used to assess the effect of plasma PLP on inflammatory and immune responses after adjusting for confounders. The results were considered statistically significant at P<0.05. Values presented in the text are means±standard deviation (s.d.).
The characteristics of the subjects in each group are shown in Table 1. Forty-three patients who met the eligibility criteria for the study were recruited. However, eight patients dropped out because of scheduling problems. The compliance of all of them was good (87.8±4.1%). Patients’ disease activity score 28 ranged from 1.7 to 6.3, indicating a moderate level of disease activity (3.2<DAS 285.1). Patients reported having a diagnosis of RA for 2.6±1.6 years. There were no significant differences between the two groups with respect to age, body mass index, blood pressures and disease activity.
Nutrient intake within the 24-h diet recall kept by subjects at baseline and week 12 was calculated (data not shown). The mean dietary intakes of vitamin B6 were 1.2±0.4 and 1.1±0.4 mg at baseline and week 12 in the control group and 1.3±0.5 and 1.3±0.4 mg at baseline and week 12 in the vitamin B6 group. The mean dietary intakes of folate were 224.3±237.3 and 257.4±183.7 μg at baseline and week 12 in the control group, and 310.5±222.2 and 293.8±197.7 μg at baseline and week 12 in the vitamin B6 group. The mean dietary intake of macronutrients, vitamin B6 and folate between the baseline and week 12 was not significantly different in each group. Total vitamin B6 and folate intakes (dietary plus supplementation) significantly correlated with plasma PLP (r=0.60, P=0.00) and serum folate (r=0.42, P=0.00) concentrations in all subjects, respectively.
The changes in vitamin B6, folate, inflammatory and immune responses before and after vitamin supplementation are shown in Table 2. At baseline (week 0), there were no significant differences in vitamin B6 and folate status, or inflammatory and immune response indicators between the two groups. After 12 weeks of vitamin supplementation, plasma PLP concentrations significantly increased in the vitamin B6 group and serum folate concentration significantly increased in both the groups. At week 0, there were no significant differences with respect to inflammatory and immune parameters between the two groups. In the group receiving vitamin B6 all patients responded, with a reduction in plasma IL-6 and TNF-α levels in response to 12 weeks of vitamin B6 supplementation. There were no significant changes in any of the immune parameters, except for the percentage of total lymphocytes in the vitamin B6 group when comparing week 0 with week 12. However, the net change in the percentage of total lymphocytes (Δ week 12−week 0) did not differ between control and vitamin B6 groups.
Linear regressions were performed to understand the effect of plasma PLP on inflammatory and immune responses (Table 3). Plasma IL-6 level remained significantly inversely affected by plasma PLP concentration after adjusting for confounders. An elevation of 1 nmol/l in plasma PLP concentration decreased the serum IL-6 level by 0.01 pg/ml. In addition, plasma PLP was positively correlated with percentage of total lymphocyte after additionally adjusting for serum folate concentration.
Our finding that vitamin B6 supplementation was ineffective in suppressing hs-CRP and ESR levels is partially in agreement with the results of a study by Chiang et al. (2005b). A possible explanation might be that the use and turnover of plasma PLP increased or redistributed PLP from plasma to erythrocyte in an acute systemic inflammatory response (Talwar et al., 2003; Gray et al., 2004; Quasim et al., 2005); plasma PLP, therefore, was negatively associated with CRP level in our earlier study, which included critically ill patients (Huang et al., 2005), and in other studies (Friso et al., 2001, 2004; Kelly et al., 2004). However, RA is a chronic rather than acute inflammatory disease; our RA patients had not only adequate but also relatively high plasma PLP concentrations before the intervention (week 0). Sufficient plasma PLP thus can be removed from the circulation to meet RA patients’ higher demands of certain tissues during inflammation. Therefore, additional vitamin B6 supplementation had no greater beneficial effect in lowering CRP or ESR levels in our patients with RA.
Pro-inflammatory cytokines such as IL-6 and TNF-α have key roles in driving the inflammation and synovial cell proliferation that characterize RA joint destruction. It is, therefore, not surprising that IL-6 and TNF-α are abundantly expressed in patients with RA (Guerne et al., 1989; Madhok et al., 1993; Choy and Panayi, 2001). As plasma PLP acts as a co-enzyme for the production of cytokines and other polypeptide mediators during the inflammatory response (Friso et al., 2001), a prolonged pro-inflammatory state may lead to plasma PLP depletion and could be expected to be inversely associated with pro-inflammatory cytokines (that is, IL-6 or TNF-α) in patients with RA. Plasma PLP levels have been shown to be inversely correlated with production of TNF-α by peripheral blood mononuclear cells from patients with RA (Roubenoff et al., 1995) and with IL-6 receptor (Gori et al., 2006). Thus, supplementation of vitamin B6 may be expected to result in decreased IL-6 or TNF-α levels in RA patients. However, 50 mg/day vitamin B6 supplementation for 30 days did not suppress pro-inflammatory cytokine production in a study by Chiang et al. (2005b). On the other hand, although vitamin B6 supplementation had no lowering effect on hs-CRP or ESR levels, 100 mg/day of vitamin B6 supplementation for 12 weeks was capable of suppressing pro-inflammatory cytokine production (that is, IL-6 and TNF-α) in our RA patients. A higher dose (100 mg/day) and longer supplementation period might be two possible reasons for the discrepancy between our findings and Chiang’s result.
Methotrexate, a cytostatic drug frequently used for the treatment of RA, could interfere with folate metabolism. Patients with RA therefore usually take folic acid supplements to compensate for the side effects of methotrexate. To eliminate the interference of folic acid, we co-intervened folic acid in this study. As was found in earlier studies (Gori et al., 2006), serum folate concentration was not associated with hs-CRP and IL-6 in our RA patients (data not shown). Beyond serum folate, plasma PLP seems to have a major role in the regulation of inflammatory response in patients with RA.
In the duration of RA disease, cytokines are implicated in each phase of the pathogenesis of RA; by promoting autoimmunity (Mclnnes and Schett, 2007), immune responses could thus be depressed in patients with RA that is characterized by a chronic inflammatory response. During the inflammatory process, T-helper cells (CD4), B-cells (CD19) and monocytes–macrophages would migrate into and remain in the synovium, which is a result of interaction of cellular adhesion molecules with counterligands expressed on extracellular matrix molecules (Fox, 1997; Moreland et al., 1997). It has been observed that serum levels of TNF-α, IL-6 and soluble CD4 were significantly higher in RA patients compared with 30 healthy subjects (Kuryliszyn-Moskal, 1998). Hussein et al. (2008) also found a high CD4+/CD8+ T-cell ratio in RA patients. Low plasma PLP concentrations have been related to alteration of immune responses, including impairment of both humoral and cell-mediated immunity (Meydani et al., 1991, 1992; Rall and Meydani, 1993). In turn, reduced plasma PLP levels may alter immunological functions. Talbott et al. (1987) indicated that increasing vitamin B6 intake could improve the immune responsiveness of both T- and B-cells. Folkers et al. (1993) treated nine healthy subjects with 300 mg/day of pyridoxine; consequently, their T4-lymphocytes and T4/T8 ratio significantly increased over 2 months. However, we only found a significant increase in the level of percentage of total lymphocytes after 12 weeks of vitamin B6 supplementation. As levels of immune parameters except for T-suppressor cell (CD8) and the mean PLP concentration were above the lower limit of normal in most of our patients, supplementation of vitamin B6 probably had no further beneficial effect in improving immune responses.
Although we calculated the sample size to meet the statistical power, the loss of eight patients from the study population might have decreased the degree of effect of plasma PLP on inflammatory and immune responses. This was a single-blind rather than a double-blind study, which may also be considered to be a limitation in this study. The third limitation in this study might be the short disease duration of our patients. Disease duration may have influenced the inflammatory and immune levels.
In conclusion, 100 mg/day of vitamin B6 supplementation suppressed pro-inflammatory cytokines (that is, IL-6 and TNF-α) in patients with RA. Our results provide valuable reference data for clinical practice with regard to the potential beneficial use of vitamin B6 to suppress inflammatory response in RA patients.
This study was supported by the National Science Council, Taiwan, ROC (NSC-94-2320-B-040-031).
About this article
Plasma pyridoxal 5′-phosphate is not correlated with hemoglobin during pyridoxine supplementation in patients with rheumatoid arthritis
European Journal of Clinical Nutrition (2011)