点击显示 收起
Institute of Clinical Chemistry and Biochemistry
University Hospital Magdeburg
Leipziger Strasse 44
D-39120 Magdeburg
Germany
E-mail: jutta.dierkes{at}medizin.uni-magdeburg.de
German Institute of Human Nutrition
Potsdam-Rehbrücke
Germany
University Hospital Hamburg-Eppendorf
Hamburg
Germany
Institute of Laboratory Medicine
Clinical Chemistry and Molecular Diagnosis
University Hospital Leipzig
Leipzig
Germany
Dear Sir:
Low plasma concentrations of pyridoxal-5'phosphate (PLP), the active metabolite of vitamin B-6, have been discussed as a cardiovascular disease (CVD) risk factor (1–3). However, this association may be caused by an influence of inflammation on plasma PLP concentrations (4) that was not corrected for in the early studies (1–3). Recently, Friso et al (5) reported in this Journal that low plasma PLP concentrations were a significant risk factor for coronary artery disease (CAD) in a case-control study in Italy. Plasma PLP was significantly and inversely associated with both angiographically defined CAD and high-sensitivity C-reactive protein (hs-CRP). However, adjustment for markers of inflammation did not change the association of plasma PLP with CAD.
The results of a retrospective case-control study of CVD risk factors in women in Germany, the Coronary Risk Factors for Atherosclerosis in Women (CORA) Study, do not confirm the results of Friso et al. The German study is a case-control study including 200 case and 255 control subjects (all women). Details of the study design were reported elsewhere (6). Cases were selected after admittance to the Department of Internal Medicine, University Hospital Hamburg–Eppendorf, for incident CAD, which was usually verified by angiography. Age-matched control subjects from the same district were invited by mailing. Control subjects with any symptoms suggesting CAD were excluded. Fasting blood samples were collected within 24 h of admittance and were analyzed for CVD risk factors including hs-CRP, homocysteine, folate, vitamin B-12, and vitamin B-6. Vitamin B-6 was measured as PLP with the use of an HPLC method with fluorescence detection.
The CORA Study used the same type of statistical analysis as was used in the study of Friso et al to ensure comparability between the 2 studies. Median PLP concentrations in the control subjects (44 nmol/L) were used to define low vitamin B-6 status. Univariate and multivariate logistic regression with log-transformed variables was performed, which allowed the analysis to be controlled for age, smoking status, hypertension, diabetes, body mass index, and concentrations of blood lipids, creatinine, hs-CRP, homocysteine, and other vitamins. Analyses were performed by using both the dichotomized PLP concentrations and quintiles of PLP.
Median PLP and hs-CRP concentrations differed significantly between case and control subjects (PLP: 30.5 and 44 nmol/L, respectively; hs-CRP: 12 and 2.2 mg/L, respectively) and were significantly associated with CAD (r = –0.37, P < 0.001). In the univariate logistic regression analysis, low PLP was significantly associated with CAD (odds ratio: 3.38; 95% CI: 2.23, 5.11). Adjustment for classic risk factors did not substantially change the association between PLP and CAD, but that association became nonsignificant after the inclusion of hs-CRP in the model (odds ratio: 1.77; 95% CI: 0.96, 3.28). This effect was seen in the analysis of both the dichotomized PLP concentrations (Table 1) and PLP quintiles (Table 2).
View this table:
TABLE 1. Odds ratios (95% CIs) for coronary artery disease according to low pyridoxal-5'phosphate (PLP) concentrations in the Coronary Risk Factors for Atherosclerosis in Women Study, 1997–20001
View this table:
TABLE 2. Odds ratios (95% CIs) for coronary artery disease according to quintiles of pyridoxal-5'phosphate concentrations in the Coronary Risk Factors for Atherosclerosis in Women Study, 1997–20001
These results contrast with those of the case-control study of Friso et al. In our study, differences in PLP and hs-CRP concentrations between case and control subjects were more pronounced, probably because of the inclusion of acute coronary syndromes, which Friso et al did not include in their study.
In the study of Friso et al, the magnitude of the association of PLP with CAD was low (odds ratio: 1.89), but it remained significant after adjustment for markers of inflammation. However, we wonder whether the results of Friso et al would have been nonsignificant, because of substantial residual confounding in their model, if they had also made an analysis by quintiles.
We suggest that the association of low PLP with CVD risk is mainly due to the effect of inflammation on plasma PLP concentrations. It is not yet known whether the measurement of whole-blood PLP concentrations would confirm this association (7). However, the results of the CORA Study and other studies of stroke (8) suggest that low plasma PLP concentrations should be evaluated in connection with markers of inflammation. Low plasma PLP concentrations in CAD patients obviously do not indicate vitamin B-6 deficiency, but they do appear to reflect systemic inflammation. A more definitive answer as to whether low PLP is associated with CVD risk independent of the inflammatory response will be obtained by studies with a prospective design.
ACKNOWLEDGMENTS
None of the authors had any personal or financial conflicts of interest with the study by Friso et al or their article in the Journal.
REFERENCES