|Home | About | Journals | Submit | Contact Us | Français|
Vitamin B6 is an important cofactor in many metabolic processes. However, vitamin B6 intake and plasma status have not been well studied in the Puerto Rican population, a group with documented health disparities.
To assess dietary intake of vitamin B6, food sources, and plasma status of pyridoxal 5′-phosphate (PLP), and their associations with plasma homocysteine in 1236 Puerto Rican adults, aged 45–75 years, living in the greater Boston area.
Baseline data were analyzed cross-sectionally.
Questionnaire data were collected by home interview. Dietary intake was assessed with a semi-quantitative food frequency questionnaire. Plasma PLP and homocysteine were assayed from blood samples collected in the home.
The mean daily intake of vitamin B6 was 2.90 ± 1.28 mg (mean ± SD) for men and 2.61 ± 1.29 mg for women (P <0.001). Approximately 11% were deficient (PLP <20 nmol/L) and another 17 % insufficient (PLP ≥20 but <30 nmol/L). Household income below the poverty threshold, physical inactivity and current smoking were significantly associated with lower plasma PLP (P <0.05). Food groups contributing most to vitamin B6 intake included ready-to-eat cereals, poultry, rice, potatoes and dried beans. However, only intake of ready-to-eat cereals and use of supplements with vitamin B6 were significantly associated with plasma PLP sufficiency (≥30 vs. <30 nmol/L, P <0.01). Both vitamin B6 intake and PLP were significantly associated with plasma total homocysteine (P < 0.001). The association between PLP and homocysteine remained statistically significant after further adjustment for plasma vitamin B12 and folate (P = 0.028).
Given the known importance of vitamin B6 to health, the high prevalence of low vitamin B6 status in this Puerto Rican population is of concern. Further work is needed to clarify the potential role that insufficient vitamin B6 may have in relation to the observed health disparities in this population.
In its physiologically active coenzyme form, pyridoxal 5′-phosphate (PLP), vitamin B6 functions in the metabolism of amino acids, nucleic acids, fats and glycogen, and the biosynthesis of heme, histamine and neurotransmitters (1). Low vitamin B6 status has been associated with weakness, sleeplessness, depression, nervous disorders, cheilosis, stomatits and impaired cell-mediated immunity (1). Low PLP concentration has also been linked with increased risk of cardiovascular disease (CVD) (2–5). During the past few decades, homocysteine has been well documented as a risk factor for CVD, although the mechanism has not been fully explored (6). Homocysteine metabolism involves two pathways: the methylation of homocysteine to methionine with folate and vitamin B12 as coenzymes, and the irreversible transsulfuration of homocysteine to cysteine with PLP as a coenzyme (7). In addition to folate and vitamin B12, higher plasma PLP has been associated with lower homocysteine (8–9). It is likely that vitamin B6 may play an even more important role in relation to plasma homocysteine in the post-folate fortification era (10).
Plasma PLP below 30 nmol/L has been used as an indicator of insufficient vitamin B6 status (11–12), while that below 20 nmol/L is thought to indicate deficiency (13). Approximately 23–27% of US adults not using supplements had plasma PLP <20 nmol/L, according to the recent National Health and Nutrition Examination Survey (NHANES 2003–2004) (14). Current recommended dietary allowance (RDA) criteria for vitamin B6 intake are mainly based on plasma PLP of at least 20 nmol/L (13). However, limited data suggest that the current RDA may not be sufficient to minimize risk of nonfatal myocardial infarction (15). In addition, the NHANES 2003–3004 showed that low plasma PLP was more prevalent in some subgroups, such as smokers and non-Hispanic blacks, even if they met the RDA for vitamin B6 intake (14). Therefore, it is important to investigate vitamin B6 intake and plasma status among understudied ethnic groups, especially among those with documented health disparities.
Puerto Ricans, the second largest Hispanic sub-group living in the United States (16), have been identified as having excess chronic health conditions including type 2 diabetes, depression and physical disability, relative to non-Hispanic whites (17–19). Limited evidence also suggests that Puerto Ricans may have lower intake and plasma status of several vitamins, including vitamin B6 (19) and vitamin B12 (20) than their non-Hispanic white counterparts.
Puerto Ricans have a unique dietary pattern compared to the overall US population (19). Rice is a major contributor to energy intake, and dried beans, rice dishes and plantains are consumed frequently in this population (21). Further, the portion size for many commonly consumed foods including rice, soups and some fruits are larger than those consumed by the general US population, while portion sizes for some vegetables tend to be smaller (21). There are few data about vitamin B6 intake and food sources of vitamin B6 in Puerto Ricans. Therefore, the aims of the present study were to: 1) characterize dietary vitamin B6 intake and food sources and plasma PLP concentration; 2) examine associations between vitamin B6 intake and plasma PLP and risk of low PLP status; and 3) evaluate associations between vitamin B6 intake, plasma PLP and plasma homocysteine among Puerto Rican adults living in the greater Boston area.
Puerto Rican adults, aged 45–75 years, were recruited from 2004 to 2009 in the greater Boston area for the Boston Puerto Rican Health Study, an ongoing longitudinal study focusing on relationships between stress, nutrition, and chronic health conditions. The study design and sampling process have been described in detail elsewhere (22–23). Briefly, blocks with ≥10 Hispanic adults were enumerated from census tracks with ≥25 Puerto Rican adults, aged 45–75 years, in the year 2000 census in the greater Boston area. One qualified participant was randomly recruited from each household identified in the enumerated blocks. Most (77.4%) of the participants were recruited this way; others were recruited through random approach at major Latino events, calls responding to flyers, or referrals from community members. We identified 2170 eligible participants. Of these, 77 were excluded because they either had serious health conditions or advanced dementia which would preclude completion of the interviews, they did not have a permanent address, or they planned to move away within two years. The remaining 2093 were invited and 1811 (86.5%) agreed to participate. Of these, 1500 (83% of those who agreed) completed the baseline interview. Most cases of those who did not complete it were due to repeated difficulties in scheduling. Those not participating in the interview tended to be male and had been in the US for a shorter time than those who completed the baseline interview (23). The sample used in our analysis consisted of 1236 participants with completed and cleaned data on both vitamin B6 intake and plasma PLP. Participants were excluded if they reported implausible energy intake (<600 kcal per day or >4800 kcal per day). Approval of the study protocol was obtained from the Institutional Review Board at Tufts Medical Center, and each participant provided written informed consent.
Data on age, sex, educational attainment, household income, acculturation, smoking, alcohol use, physical activity, history of diseases and medication use, and anthropometric measures were collected during interviews in the home (23). Physical activity was estimated based on a modified Paffenbarger questionnaire of the Harvard Alumni Activity Survey (24). Physical activity score was constructed by weighting time spent in various activities by factors that parallel oxygen consumption rates associated with varying types of physical activity.
Educational attainment was dichotomized as ≥9th or <9th grade education. The poverty ratio (household income relative to a family's appropriate poverty threshold) was calculated with information on self-reported household income. The ratio was dichotomized as being below poverty (yes) or not, based on eligibility for food assistance programs (i.e., 120% of the poverty levels for household size and year of data collection). An acculturation score was calculated, based on answers to questions to assess the extent of use of English and/or Spanish: at work, to watch television, to listen to the radio, to read newspapers/books, to speak with neighbors, to talk with friends and to talk with family members. The score ranges from 0 (only using Spanish) to 100 (fully acculturated, only using English) (25). Smoking status was categorized as never smoking (<100 cigarettes in entire life), former smoking or current smoking. Alcohol use was classified as not current, current moderate (≤ 1 drink daily for women or ≤ 2 drinks daily for men), or current heavy (>1 drink daily for women and >2 drinks daily for men). Physical activity scores were dichotomized as <30 (physically inactive) or ≥30 (physically active). Obesity status was classified as normal weight (BMI <25 kg/m2), overweight (BMI ≥25 and <30 kg/m2) or obese (BMI ≥30 kg/m2). Diabetes was defined as `yes' if a participant had fasting glucose ≥7.0 mmol/L or was taking medication for diabetes (26).
Dietary intake of vitamin B6 was estimated from a semi-quantitative food-frequency questionnaire (FFQ) specifically designed for this population (21). This questionnaire has been validated against several measures, including plasma carotenoids (27), and vitamin B12 (20) in Hispanic adults, aged 60 years and older. The nutrient database for this FFQ (21) was developed using the Nutrition Data System for Research (NDS-R) software (Nutrition Coordinating Center, University of Minnesota, Minneapolis, MN, Food and Nutrient Database, release 2007).
Overnight fasting blood samples (12 h) were drawn by a certified phlebotomist in the home. Aliquots were saved and stored at −80°C until processed. Total plasma homocysteine was determined by high-performance liquid chromatography, as described previously (CV = 7.8%) (28). Serum creatinine was detected with a colorimetric, kinetic reaction on the Olympus AU400e with Olympus Creatinine Reagents (OSR6178) (Olympus America Inc., Melville, NY). Plasma PLP was determined enzymatically using tyrosine decarboxylase (CV = 16%) (29).
All statistical analyses were conducted with SAS 9.1.3 (SAS Institute Inc, Cary, NC); and all tests were 2-sided with P <0.05 considered statistically significant. Skewed variables, including PLP and total homocysteine, were natural log-transformed to normalize distributions before analysis. General linear models were used to estimate associations with sex and other continuous dependent variables, and logistic regression models were used for dichotomized dependent variables. Adjusted means of plasma PLP were calculated and plotted against quintile medians of vitamin B6 intake for the full sample, and for those not using supplements with vitamin B6. Medians of each quintile of vitamin B6 intake were used for trend tests. Similar statistical analyses were conducted for plasma total homocysteine against vitamin B6 intake and plasma PLP. Covariates included age, sex, energy intake, protein intake, educational attainment, poverty status, acculturation score, smoking status, alcohol use, physical inactivity, obesity status, presence of diabetes, and plasma creatine. Associations of plasma total homocysteine with vitamin B6 intake and Plasma PLP were further adjusted for plasma folate and vitamin B12.
To compare food sources of vitamin B6 intake in this population with general U.S. adults, food groups were constructed based on USDA classifications (30), with slight modification to accommodate Hispanic foods. For example, plantains and non-potato root crops, like cassava, were added (21). Food sources of vitamin B6 intake were identified and ranked for the whole sample. In addition, major vitamin B6 sources were identified and ranked for participants above and below the 20 and 30-nmol/L cutoff points for plasma PLP. Adjusted plasma PLP means for those using versus not using supplements with vitamin B6 (yes, or no) were calculated with general linear models, after adjustment for covariates. Prevalence of, and odds ratios for, low plasma PLP status by vitamin B6 supplement use status were compared and calculated with logistic regression models, after adjustment for covariates. Similar analyses were performed to examine plasma PLP status by tertile categories of ready-to-eat cereal intake.
The mean age of the sample was 57.3 (SD: 7.6) years. Women had significantly lower intake of energy, protein and dietary vitamin B6, were more likely to not meet the RDA for vitamin B6 intake, and had lower plasma total homocysteine than men (Table 1). There were no significant differences in ratios of total vitamin B6 to energy or protein intake, plasma PLP concentration, or prevalence of inadequate or insufficient vitamin B6 status between women and men. Approximately 11 % had deficient plasma PLP (<20 nmol/L), and another 17% had insufficient plasma PLP (≥20, <30 nmol/L) (Table 2).
There were no significant differences in total vitamin B6 intake, plasma PLP concentration, prevalence of plasma PLP deficiency (<20 nmol/L) or insufficiency (<30 nmol/L) by age or educational attainment (Table 2). However, participants living in poverty were significantly less likely to use supplements with vitamin B6, and had lower plasma PLP concentration and higher prevalence of plasma PLP deficiency and insufficiency than those with higher income. More acculturated participants had significantly higher intake of total vitamin B6 and were more likely to use supplements. Current smokers had lower plasma PLP concentration and higher prevalence of plasma PLP deficiency and insufficiency, despite similar vitamin B6 intake, when compared to non-smokers. In contrast, moderate, but not heavy, alcohol drinkers had higher plasma PLP and lower prevalence of deficiency and insufficiency, when compared to non-drinkers.
Significant and similar Pearson partial correlation coefficients were observed between log-transformed plasma PLP and total vitamin B6 intake (r = 0.27) (Figure 1), vitamin B6 to energy intake ratio (r = 0.23) and vitamin B6 to protein intake ratio (r = 0.23), after adjustment for covariates (all P <0.001). Generally weaker, but also significant and similar correlation coefficients were observed among non-supplement users (r = 0.13, 0.16, 0.16, respectively, all P <0.001).
In addition to vitamin B6 supplements, a range of foods contributed to vitamin B6 intake in this population (Table 3). Ready-to-eat cereals were the major contributor (10.0%) among food sources, due to fortification with vitamin B6 and frequent consumption (52% reported intake at least once per week). Poultry ranked second (9.6%) due to high vitamin B6 content (0.41 mg/100g) and high intake (mean = 52 g/d). Although rice is not considered as a good source of vitamin B6 (0.10 mg/100g), it was the third contributor (6.5%) due to high intake (mean = 154 g/d) by most participants. Potatoes ranked fourth (6.3%) as a good source (0.35 mg/100g) with high intake (mean = 42 g/d), and dried beans (0.19 mg/100 g) ranked fifth (5.1%).
To better understand sources of vitamin B6 intake contributing to plasma PLP concentration, sources for participants above and below the 20- and 30-nmol/L cutoff points, respectively, were compared (Table 4). As expected, participants with PLP ≥30 nmol/L had significantly higher vitamin B6 intake from supplements than their counterparts with lower PLP concentration. Participants with plasma PLP ≥30 nmol/L also had higher vitamin B6 intake from ready-to-eat cereals and poultry, but had lower vitamin B6 from potatoes than those with PLP <30 nmol/L. Participants with plasma PLP ≥20 nmol/L had higher vitamin B6 intake from supplements, and from milk, compared with those with plasma PLP <20 nmol/L.
Use of supplements with vitamin B6 and intake of ready-to-eat cereals were significantly associated with higher plasma PLP concentration and lower prevalence of low plasma PLP status (Table 5). Vitamin B6 supplement users had 43% lower odds (odds ratio, 95% confidence interval [CI]: 0.57, 0.39–0.83) of plasma PLP < 30 nmol/L than non-supplement users. Among non-supplement users, those in the upper tertile of ready-to-eat cereals had almost 50% lower odds for having plasma <30 nmol/L (odds ratio, 95% CI: 0.48, 0.32–0.71) or for having plasma <20 nmol/L (odds ratio, 95% CI: 0.54, 0.29–0.99) than those in the lowest tertile.
As expected, higher vitamin B6 intake was associated with lower plasma total homocysteine after adjustment for covariates (P for trend < 0.001) (Figure 3.A). A similar trend was observed between plasma PLP and homocysteine (Figure 3.B). The mean adjusted total homocysteine concentration in the lowest quintile of PLP was 9.84 μmol/L (95% CI: 9.33, 10.39 μmol/L) relative to 8.83 μmol/L (95% CI: 8.39, 9.29 μmol/L) in the highest quintile. Further adjustment for dietary and plasma folate and vitamin B12, respectively, attenuated the associations between plasma homocysteine and vitamin B6 intake and plasma PLP, although the latter remained statistically significant (P for trend = 0.028).
In this population of Puerto Rican adults aged 45–75 years, lower proportions of participants had intakes below the RDA for dietary vitamin B6 intake (22% of women, 19% of men) than observed in our previous study of Puerto Ricans and Dominicans aged 60 years and older (39%), in which the same food frequency questionnaire was used (19). However, similar proportions in both studies had deficient or insufficient plasma PLP status. The prevalence of vitamin B6 deficiency (PLP <20 nmol/L) was comparable with that of the NHANES 2003–2004 (6%~7%) among supplement users, but was lower among non-supplement users (23%~24%) in a similar age range (14). Our subgroup analyses showed that plasma PLP concentration was significantly lower among participants living in poverty, likely due to the lower use of supplements with vitamin B6, a major contributor to higher plasma PLP concentration (≥30 nmol/L).
Acculturation has been associated with diet changes among Latinos living in the US (31). In this study, more acculturated participants had higher total intake of vitamin B6 than those less acculturated. In contrast, results from the NHANES III (32) showed that Mexican Americans born in the US had lower intake of vitamin B6 than those born in Mexico, regardless of language use. This difference may be due to differences in the traditional dietary patterns of Mexican Americans relative to Puerto Ricans, as well as to different dietary responses to acculturation. Available data suggest that the acculturation process may have varied effects on dietary change (31). Longitudinal analyses are needed to fully examine the effects of acculturation on dietary intake.
Consistent with several previous studies (14, 33–34), smokers were more likely to have low plasma PLP despite similar vitamin B6 intake. Although the mechanisms are not fully understood, smokers have elevated heat-stable alkaline phosphatase (ALP) isoenzyme, which may have a higher affinity for PLP than other ALP isoenzymes, resulting in increased hydrolysis of PLP (34). In addition, smoking has been positively associated with urinary albumin excretion (35). Because PLP binds to albumin in circulation, lower albumin may result in more free PLP which may, in turn, be hydrolyzed by ALP (36).
Interestingly, moderate alcohol intake was associated with higher plasma PLP concentration. This association remained unchanged after further adjustment for total vitamin B6 intake or use of supplements with vitamin B6. Consistently, several studies have demonstrated that intake of alcoholic beverages was positively associated with plasma PLP concentration (37–39). It has been suggested that the bioavailability of vitamin B6 from alcoholic beverages, such as beer, may be relatively high (40). In addition, alcohol may increase plasma PLP concentration by shifting homocysteine metabolism from the transsulfuration to the transmethylation pathway (38–39). Moderate alcohol intake has been shown to have cardioprotective effects (41). Because lower plasma PLP concentration has been associated with higher incidence of CVD (2–5), further studies are warranted to test whether associations of moderate alcohol intake with CVD is mediated by plasma PLP.
Due to the unique dietary pattern of Puerto Rican adults (19, 21), the dietary sources of vitamin B6 intake differed somewhat from sources for the general US adults. Cotton et al reported the top five food sources for vitamin B6 in the USDA's 1994 to 1996 Continuing Survey of Food Intakes by Individuals (CSFII) as ready-to-eat cereals (13.0%), poultry (9.0%), beef (8.6%), potatoes (8.5%), and bananas (5.8%) (30). In the current sample of Puerto Ricans, ready-to-eat cereals and poultry were also the first two major food sources of vitamin B6. However, these were followed by rice, potatoes and dried beans, while beef ranked sixth.
As expected, higher intakes of vitamin B6 from supplements and ready-to-eat cereal were significantly associated with higher plasma PLP concentration and lower prevalence of plasma PLP <30 nmol/L. Previous studies have shown that vitamin B6 from oral supplements can be absorbed quickly, resulting in significant increases in plasma PLP (42). Many ready-to-eat cereals have been fortified with vitamin B6, and these have been shown to improve plasma PLP concentration (10, 43).
Higher dietary intake of vitamin B6 was inversely associated with plasma homocysteine in this study. However, the association became non-significant after further adjustment for dietary intake of folate and vitamin B12. It is likely that overlapping dietary sources contribute to vitamin B6 and folate intake, especially dietary supplements. The association between plasma PLP and homocysteine was also reduced when folate and vitamin B12 were controlled, but it remained statistically significant. This is consistent with several studies showing inverse associations of fasting plasma PLP with plasma total homocysteine (8–9) and/or lowering effects of vitamin B6 supplements on homocysteine (10), although not all have shown this (42).
The high proportion of participants with plasma PLP <30 nmol/L in this study suggests that vitamin B6 intake and status need more focused attention in Puerto Rican adults. Previous studies have demonstrated that even “adequate” vitamin B6 intake does not always result in adequate plasma status (14). In the current study, 10.0% and 26.4% of these who met the RDA for total vitamin B6 intake had plasma PLP <20 nmol/L and <30 nmol/L, respectively. Therefore, intake of vitamin B6 above the current RDA may be advisable. To date, there is no consensus on whether increasing intake of vitamin B6 from food or supplements is more beneficial for older adults relative to younger adults (14). In this sample of Puerto Rican adults, only supplement use clearly differentiated those with vitamin B6 deficiency (PLP <20 nmol/L) versus those without deficiency. This agrees with results from the NHANES 2003–2004: only 6% of supplement users aged 65 years and over had vitamin B6 deficiency whereas 24% among non-supplement users in the same age group; similar prevalence was observed for those aged 45–64 years (7% and 23%, respectively) (14).
There were some limitations in our study. Dietary intake was measured by food frequency questionnaire, which is only semi-quantitative. However, this food frequency questionnaire was developed and validated for this population. Secondly, we did not measure erythrocyte PLP. Erythrocyte PLP has been suggested as a better indicator of vitamin B6 status among smokers and individuals with illness (34, 44). However, plasma PLP is the most commonly used measure of vitamin B6 status, allowing us to compare against other populations.
Middle-aged and older Puerto Rican adults living in the greater Boston area had high prevalence of inadequate B6 status, particularly among those with household incomes below poverty, those who were physically inactive, and current smokers. Further, vitamin B6 status was associated with plasma homocysteine concentration. Although a variety of food groups contributed to vitamin B6 intake, supplements containing vitamin B6 and ready-to-eat cereals fortified with vitamin B6 appeared to be most protective. More research is needed to clarify the role of low vitamin B6 in relation to observed health disparities in this population, and to design effective interventions to improve vitamin B6 status.
Xingwang Ye, PhD Current affiliation: Postdoctoral Associate Department of Health Sciences 312 Robinson Hall, Northeastern University 360 Huntington Avenue, Boston, MA 02115 Tel: 617-373-4293; Fax: 617-373-2968, xi.ye/at/neu.edu
Janice E. Maras, MS Current affiliation: Research Manager Department of Health Sciences 312 Robinson Hall, Northeastern University 360 Huntington Avenue, Boston, MA 02115 Tel: 617-373-3665, Fax: 617-373-2968, j.maras/at/neu.edu
Katherine L. Tucker, PhD Current affiliation: Professor and Chair Department of Health Sciences 316 Robinson Hall Northeastern University 360 Huntington Avenue Boston, MA 02115 Phone: 617-363-3666, Fax: 617-373-2968, kl.tucker/at/neu.edu
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.