Specificity of the assay was determined using three different methods. The cross-reactivities of steroids structurally related to testosterone (along with the bile acids found in feces that contain the same cholesterol backbone) were assessed for the antibody. The percentage cross-reactivity was defined as 100X/Y where X is picograms of testosterone and Y is picograms of steroid required to produce 50% inhibition of HRP-testosterone hapten binding to antibody. Parallel studies were done by taking feces and extracting them two times with 3.8 ml of ethyl ether as described below. The fecal extracts were serially diluted and compared to standards for parallelism over the range of 12-96 pg of testosterone (n = 4). In a final determination, pooled fecal samples were extracted with ether, and the extract was separated on HPLC by methods described later. Fractions were analyzed using the ELISA. antibiotic levaquin
Sensitivity was calculated from 95% confidence limits at the zero point of the standard curve. The interassay coefficient of variation was determined using standards of testosterone at low, medium, and high concentrations (20 plates each with sample run in duplicate). The intraassay coefficient of variation was determined using pooled samples of fecal extract at low, medium, and high concentrations of testosterone (24 aliquots of each sample run on two separate plates). Student’s /-tests were used to test the significance of the difference between slopes under the assumption that the standard errors of each of the populations were equal. Confidence limits were obtained using the computer spreadsheet Excel (Microsoft, Redmond, WA).