Table 2.

Multiple linear regression models relating vitamin intake and markers of inflammation to circulating vitamin B₆ concentrations.^a

	Log (vitamin B6 in nmol/L)
	Model 1b		Model 2c
Independent variables	b (SE)	P	b (SE)	P
Age (years × 10)	−0.042 (0.006)	<0.0001	−0.036 (0.006)	<0.0001
Square root of alcohol intake	0.012 (0.004)	0.001	0.013 (0.004)	0.001
Log(β-carotene intake in μg/day)	0.194 (0.048)	<0.0001	0.191 (0.048)	<0.0001
Log(vitamin B6 intake in mg/day)	0.256 (0.126)	0.042	0.220 (0.126)	0.079
Log(α-tocopherol in μmol/L)	0.296 (0.071)	<0.0001	0.258 (0.071)	<0.0001
Log(folate in nmol/L)	0.132 (0.036)	<0.0001	0.108 (0.037)	0.004
Log(vitamin B12 in pmol/L)	0.109 (0.035)	0.002	0.083 (0.036)	0.022
Albumin (mg/dL)	0.180 (0.029)	<0.0001	0.179 (0.029)	<0.0001
Log(CRP in mg/L)	−0.116 (0.020)	<0.0001	−0.123 (0.018)	<0.0001
Log(IL-6r in ng/L)	−0.119 (0.036)	0.001	−0.117 (0.036)	0.001

^aIn the multiple linear regression analysis, circulating vitamin B₆ (serum concentrations) was used as the dependent variable and age, sex, serum creatinine, serum albumin, smoking habits, total energy intake, inflammatory markers, nutrient intakes, plasma Hcy concentrations, and serum vitamin concentrations were used as covariates. In the final regression model, all variables not significantly and independently associated with serum vitamin B₆ concentrations were removed through backward selection method.

^bModel 1 included age; sex; serum creatinine and albumin; total energy intake; smoking habits; folate, vitamin B₆, retinol, β-carotene, vitamin C, vitamin E, and alcohol intakes; circulating concentrations of vitamin B₁₂, folate, α-tocopherol, γ-tocopherol, CRP, IL-6, IL-1 receptor agonist, and IL-6r; and leukocyte count as covariates.

^cModel 2 included the same covariates as model 1 plus plasma Hcy concentrations.