A theoretically sound and practicable equilibrium dialysis method for measuring percentage of free testosterone.

Free testosterone measured in serum equilibrated in vitro is considered a good index of biologically available testosterone even though a large part of free testosterone in vivo is derived locally from rapid dissociation of testosterone bound to albumin. The most accurate method for measuring free testosterone, however, is unsettled. The classical method--equilibrium dialysis--has been questioned because of the dilution of serum that it entails and the previous inability to achieve identical results with diluted and undiluted serum. Essentially identical measurements of free testosterone were achieved in diluted and undiluted charcoal-stripped serum by using the dialysis method and calculation reported here. The measured free testosterone in undiluted whole serum from women was only 4-6% lower than the estimated physiological values. These results were obtained using a validated calculation, controlling pH, using physiological bicarbonate buffer at 37 degrees C, maintaining a constant free ligand concentration for dilutions, measuring the water gain by the dialysis bag, and using highly purified labeled testosterone. The mean free testosterone for normal women was 0.17 ng/dl (0.11-0.23) and for hirsute women was 0.49 ng/dl (0.27-0.71). The testosterone not bound to testosterone-estradiol binding globulin, calculated from free testosterone and albumin concentrations, was close to the production rate/min of testosterone. The method should be adaptable to other ligands.