点击显示 收起
Department of Clinical and Experimental Medicine
University of Verona School of Medicine
Policlinico GB Rossi
P. le L.A. Scuro 10
37134 Verona
Italy
E-mail: simonetta.friso{at}univr.it
Dear Sir:
We thank Dierkes et al for their interest in our recent report on low vitamin B-6 concentrations and coronary artery disease (CAD) risk (1). In an earlier case-control study of angiographically defined CAD patients, we extended findings from a previous population-based study (2) and observed that low plasma pyridoxal-5'phosphate (PLP) concentrations are inversely related to major markers of inflammation and independently associated with increased CAD risk (1). Such an association of low PLP concentrations with higher risk of CAD remained significant even after the inclusion, in a multivariate logistic regression model, of high-sensitivity C-reactive protein (hs-CRP), fibrinogen, and variables related to homocysteine metabolism, such as total plasma homocysteine, vitamin B-12, and folate (1).
Several differences between our study and that of Dierkes et al may account for the dissimilar results. Their study, the Coronary Risk Factors for Atherosclerosis in Women (CORA) Study (3), was conducted only among women, whereas both sexes were represented in our study, and, in fact, there was a greater number of men than of women in our study. The participants in the CORA Study were selected for recent acute coronary symptoms, but that condition was an exclusion criterion in our study. Precisely because of the potential confounding role of inflammatory markers in the risk related to low PLP in CAD patients, the enrollment of women in acute phase may account not only for the higher hs-CRP in cases than in control subjects in the CORA Study but also for the lack of a significant association between low PLP and CAD after adjustment for hs-CRP, which was at a borderline concentration (odds ratio: 1.77; 95% CI: 0.96, 3.28). Moreover, female control subjects were enrolled only on the basis of the absence of clinical symptoms and not on that of angiography documentation. The inclusion of asymptomatic subjects who potentially had coronary atherosclerosis may also have influenced the significance of the association in the CORA Study. Indeed, it should be taken into account that the influence of low PLP on the estimate of risk in a multifactorial disease such as CAD is slight but significant [odds ratio: 1.89; 95% CI: 1.18, 3.03; P = 0.008 (multivariate logistic regression)] (1). This observation is substantiated in our study by a threshold effect for CAD risk-related PLP concentrations corresponding to the 50th percentile, an effect that is most likely due to the influence of PLP on CAD risk at the lowest concentrations.
Dierkes et al also performed a dichotomized analysis for the association with low PLP that took into account the median concentrations of PLP among control subjects in their study; this followed our model of analysis. The value that they observed, however, was higher than that in our study (44 and 36.3 nmol/L, respectively). Certainly, neither of those values can be regarded as a clear vitamin deficiency, but, considering that even a moderate vitamin B-6 impairment may influence the estimate of CAD risk, as shown by the results of our study, such a difference is likely to explain the dissimilar findings between the CORA Study and ours. Furthermore, Dierkes et al observed that the association of PLP with CAD risk became nonsignificant after the inclusion of hs-CRP in an analysis of PLP divided into quintiles, and they speculated that the association of low PLP with CAD is mainly due to the effect of inflammation on plasma PLP concentrations. As stated in our report (1), preliminary analyses by likelihood-ratio tests excluded any significant interaction between PLP and hs-CRP in our study, whether evaluated on the basis of the 50th percentile (chi-square test: 0.03, P = 0.08605) or of quintiles (chi-square test: 0.13, P = 0.7190). This finding further defined the independent association of low PLP with CAD risk and showed an additive effect of low PLP to that conferred by high hs-CRP, which was seen in the progressive increase in the estimate of CAD risk across increasing hs-CRP quintiles (1). Unfortunately, the data so far available from the literature do not allow a definite conclusion as to whether low PLP reflects systemic inflammation. In a study by Kelly et al (4), low PLP status was associated with stroke and only partially mediated via inflammation, as expressed by the major marker hs-CRP (4). Moreover, we can assume that, considering the median value of PLP in the CORA Study, even the lowest PLP quintile may be still within a range of values that may not allow detection of an influence on the risk of disease. We expect that the analysis by quintiles among the 200 cases and 255 controls of the CORA Study did not provide enough statistical power.
We certainly concur with Dierkes et al that further studies, particularly those with a prospective design, will better clarify the role of low PLP in CAD. It will be important to assess more precisely both the relation between PLP and inflammation and their specific contribution to the mechanisms underlying atherogenesis so that the role of this vitamin in the modulation of CAD risk can be elucidated.
ACKNOWLEDGMENTS
None of the authors had any personal or financial conflict of interest with regard to this work.
REFERENCES