Dietary intake and/or circulating concentrations of
vitamin B6 have been associated with risk of
cancer, but results are inconsistent and mechanisms uncertain.
Pyridoxal 5'-phosphate (PLP) is the most commonly used marker of B6 status. We recently proposed the ratio 3-hydroxykynurenine/
xanthurenic acid (HK/XA) as an
indicator of functional
vitamin B6 status, and the
4-pyridoxic acid (PA) /(
pyridoxal (PL) +PLP) ratio (PAr) as a marker of
vitamin B6 catabolism during
inflammation. We compared plasma PLP, HK/XA and PAr as predictors of
cancer incidence in a prospective community-based cohort in Norway. This study included 6,539 adults without known
cancer at baseline (1998-99) from the Hordaland Health Study (HUSK). HR and 95% CI were calculated for the risk of overall and site-specific
cancers using multivariate Cox proportional hazards regression with adjustment for potential confounders. After a median follow-up time of 11.9 years, 963
cancer cases (501 men and 462 women) were identified. Multivariate-adjusted Cox-regression showed no significant relation of plasma PLP or HK/XA with risk of incident
cancer. In contrast, PAr was significantly associated with risk of
cancer with HR (95% CI) = 1.31 (1.12-1.52) per two standard deviation (SD) increment (p < 0.01). Further analysis showed that PAr was a particular strong predictor of
lung cancer with HR (95% CI) = 2.46 (1.49-4.05) per two SD increment (p < 0.01). The present results indicate that associations of
vitamin B6 with
cancer may be related to increased catabolism of
vitamin B6, in particular for
lung cancer where
inflammation may be largely involved in
carcinogenesis.