Validity of radioimmunological methods for determining free testosterone in serum

Objective. To assess serum free testosterone (free T) concentrations in var ious groups of women by means of an ultrafiltration technique (UFT) or two RIAs and to investigate the influence of levonorgestrel (LNG) and norethist erone (NET) on free T. Design. Experimental study with serum samples from various groups of patien ts. Setting: Obstetrics and gynecology department of a university. Patient(s): Samples from 20 normal healthy women, 20 menopausal women, 21 p regnant women, 23 women using oral contraceptives (OC), 20 hyperandrogenemi c women, 20 hirsute patients with normal total testosterone (total T), 18 p atients with acne (normal total T) and 9 women with adrenogenital syndrome (AGS). Intervention(s): Blood samples were taken by venipuncture. Main outcome Measure(s): Serum concentrations of free T. Result(s): Mean fr ee-T levels determined by UFT amounted to 5-6 pg/mL in most groups, except in patients with hyperandrogenemia (8.3 pg/mL) or AGS (10.7 pg/mL) and in w omen using OC (1.9 pg/mL). The values were twofold to fourfold and threefol d to sevenfold higher than those determined with the RIAs. The proportion o f free T as determined by UFT was 0.9-1.2% of total T and was 0.55% during OC treatment. The presence of 10 ng/mL of LNG or NET increased free-T level s by only 13%. Conclusion(s): Serum free-T levels determined with RIA are much lower than those measured with the UFT. The ratio between the UFT and RIA values may v ary depending on the endocrine status. In women with normal or elevated sex hormone-binding globulin levels, nortestosterone derivatives increase free T levels only slightly. (Fertil Steril(R) 2001,76:576-82. (C) 2001 by Amer ican Society for Reproductive Medicine).