Progesterone in saliva: pitfalls and consequent implications for accuracy of the determination

The concentration of steroid hormones in saliva is believed to reflect the concentration of free hormone in blood. Because the assay for progesterone in saliva has not been rigorously validated, we investigated some of the analytical variables involved. Saliva samples were divided into two portions. One was centrifuged and the supernate used for extraction; the other was homogenized by sonication and used as such for extraction. Progesterone concentrations in homogenized whole saliva were double or triple those in supernates. By equilibrium dialysis we established that 85% of progesterone was in the free form in supernates but only 60 to 70% in homogenized whole saliva, depending upon the phase of the cycle of the subjects. Salivary flow was stimulated by crystals of citric acid; unstimulated and stimulated saliva samples were obtained from the same person. Progesterone concentrations were significantly (p less than 0.02) higher in stimulated samples than in unstimulated ones, particularly in samples collected during the luteal phase.