• plasma (7-73 μl) was made up to 100 μl with assay buffer
• radioactive testosterone (2000 dpm in 10 μl assay buffer) was added and the samples stored at 4 °C overnight
• samples were extracted twice with redistilled dichloromethane (3 and 2 ml)
• the extract was blown down under a stream of nitrogen at 45 °C
• samples were loaded onto celite columns (Wingfield & Farner 1975) in two steps with 2% redistilled ethylacetate in isooctane (volume per volume; 2 x 250 μl)
• columns were washed with 4 ml isooctane and 1.5 ml of 10 % ethylacetate in isooctane (v/v)
• testosterone was eluted from the columns with 2 ml of 20 % ethylacetate in isooctane (v/v)
• eluates were blown down under a stream of nitrogen at 45 °C
• samples were reconstituted in 150 μl assay buffer
• recovery of testosterone was assessed in 40 μl of the sample (average recovery = 90%)
• 30-100 μl of the sample was measured for T in the RIA
• RIA procedures were as described by Fusani et al. (2000)
• sensitivity of the assay was 0.30 pg per tube, intra assay variation was 2.8%, and inter assay variation was 7.6%

Back