Dietary and other determinants of vitamin B6 parameters

Objective: To assess the dietary, physical, biochemical and lifestyle determinants of vitamin B6 status parameters among healthy adults. Design and subjects: Dietary intake and status variable as well as other relevant characteristics were determined among 444 adults, aged 20-79 y and stratified for sex and 10-years age classes with a randomly selected control group (n = 300) and an over representation of persons with a low habitual vitamin B6 intake (n = 144). Results and conclusions: The direct status parameters (plasma pyridoxal-5'-phosphate (PLP), plasma pyridoxal + PLP (PL + PLP), and 4-pyridoxic acid (4-PA) excretion) were clearly related to dietary variable and plasma concentrations of vitamin C (women only), vitamin B12 and folate. The total percentage of variance in p-PLP explained in multivariate regression analysis was 41 and 30% in men and women, respectively. The most important explaining variables besides vitamin B6 intake were variable closely related to PLP-metabolism: albumin and alkaline phosphatase. Biochemical function related status parameters showed less statistically significant correlations with dietary variable. Four to 24% of the variance in the stimulation coefficients of erythrocyte aminotransferase (α-EAST) and erythrocyte alanin aminotransferase (α-EALT) and change in homocysteine excretion after a methionin load was explained by a combination of dietary, physiological and lifestyle related variables. The low percentages explained for some variable, notably α-EAST (women) and the change in homocysteine excretion after a methionin load (men), suggests that these parameters are not sensitive to the level of vitamin B6 intake as found in the present study. Alcohol contributed in many of the explaning models for vitamin B6 status parameters. Therefore, when assessing the vitamin B6 status of a population, it is important to estimate the level of alcohol intake as well. Furthermore, the results illustrate the importance of sex related differences in the metabolism of some parameters, especially homocysteine excretion. The variance in the clinical function related status parameter measured in the present study, handgripstrength, was explained for 50% by a combination of age, body weight and Quetelet Index (QI) with no important contribution of dietary variable.