RESUMO
BACKGROUND: A high-performance liquid chromatographic (HPLC) assay for vitamin B6 in human serum was compared with a novel microbiological assay (ID-Vit) that uses microtitre plates precoated with a specific microorganism, thus avoiding numerous problems associated with the use of stock cultures utilized by common other microbial assay mit B6. METHODS: Data obtained using HPLC were compared with 1D-Vit results in 170 healthy individuals and in 68 patients with coronary artery disease (CAD, 37 with acute coronary syndrome [ACS], 31 with stable CAD). Regression and Bland-Altman analysis were performed. Homocysteine in CAD patients was measured by HPLC. RESULTS: The ID-Vit assay correlated well with the HPLC assay (Pearson's r = 0.89 [p < 0.0001] in healthy and 0.82 [p < 0.001] in CAD individuals). Bland-Altman analyses revealed good agreement between the results of both methods in both cohorts, with > or = 95% of all values grouping within the lines of agreement. In CAD patients, mean homocysteine values did not differ between stable CAD and ACS and were normal. Thirty-seven percent of CAD patients had estimated glomerular filtration rates (GFR) below 60 mL/min/1.73m2. GFR correlated inversely with homocysteine levels (r = -0.80, p < 0.001) whereas neither HPLC nor ID-Vit values for B6 did. CONCLUSIONS: ID-Vit assay and the HPLC standard are in very good agreement. The new assay can easily be automated and is less laborious than common microbiological assays. The lack of correlation between B6 vitamin and homocysteine can be accounted for by the fact that mean vitamin B6 in our CAD patients was in the normal range and that a relevant percentage of patients had chronic renal disease.