Comparison of plasma and urinary levels of 2-hydroxyestrogen and 16α-hydroxyestrogen metabolites

A modified ELISA assay for measurement of the two estrogen metabolites 2-hydroxyestrone (2OHE1) and 16 alpha-hydroxyestrone (16 alpha OHE1) in plasma and serum has been developed. Previously, these have only been measured in urine. It is not known how well the measurements of these metabolites in urine and plasma are correlated. The goal of this study was to compare urinary and plasma levels of 2OHE1 and 16 alpha OHE1 and their ratios and to explore how they were affected by ethnicity, dietary and genetic factors, and medication use. Blood and urine samples were obtained from 511 nulliparous women, aged 17-35, from four ethnic groups during the same visit at the study center, on a random day of the menstrual cycle. The overall correlation between the 2OHE1/16 alpha OHE1 ratio in plasma and urine was fair (r(s) = 0.52; p < 0.0001). In general, the correlation between the 2OHE1/16 alpha OHE1 ratio in urine and plasma was higher among women not using oral contraceptives (OCs) (r(s) = 0.58; p < 0.0001) than among women currently using OCs (r(s) = 0.34; p < 0.0001). The correlation was highest for samples obtained during the mid-cycle in among non-OC users (r(s) = 0.83; p < 0.0001). Among non-OC users, the urinary 2OHE1/160 alpha OHE1 ratio was stable over the menstrual cycle while there was an increase in the plasma 2OHE1/16 alpha OHE1 ratio. The strongest factors predicting discordance between the urinary and plasma 2OHE1/16 alpha OHE1 ratios among non-OC users were a baseline urinary 20HE1/16 alpha OHE1 ratio in the three upper quartiles (p < 0.001), the menstrual cycle phase (p = 0.001), and the number of cups of coffee consumed per day (p = 0.006). Among current OC users, the strongest predictors of discordance between the urinary and plasma 2OHE1/16 alpha OHE1 ratios were a baseline urinary 2IHE1/16 alpha OHE1 ratio in the three lower quartiles (p < 0.001), being black (p = 0.001), and being Asian (p = 0.014). In conclusion, we found that the correlation between the two methods was fair and varied according to the baseline urinary 2OHE1/16 alpha OHE1 ratio, ethnic group, OC status, coffee consumption, and time of menstrual cycle when the samples were obtained. (C) 2005 Elsevier Inc. All rights reserved.