Neurosteroid interactions with GABA A receptors are not completely understood. The GABA A receptor is a pentameric complex with at least 21 subunit types involved in receptor assembly: 1-6 , ß 1-4 , 1-4 , , , , , 1-3 (Bonnert et al., 1999; Bormann, 2000; Penschuck et al., 1999). As a result, more than 500,000 different GABA A receptor subunit combinations are theoretically possible (Lambert et al., 1996). Adding yet another level of molecular complexity, subunit composition appears to determine the pharmacological profile of GABA A receptors. For example, benzodiazepines bind to GABA A receptors but require the presence of a 2 subunit to potentiate GABAergic neurotransmission (Pritchett et al., 1989), suggesting that benzodiazepine activity may be limited to a specific subset of GABA A receptors. In contrast, the neurosteroid allopregnanolone acts at a wide variety of GABA A receptor subtypes with similar potency and efficacy and does not exhibit stringent GABA A subunit specificity, possibly reflecting a broad spectrum of action in the central nervous system (Puia et al., 1990).
The course of GAD tends to be chronic in primary care patients, and GAD may "switch" to other diagnoses, particularly depression and somatoform disorders [14, 15] . GAD is associated with impairments in psychosocial functioning, role functioning, work productivity, and health-related quality of life comparable to major depressive disorder or panic disorder. Patients with GAD and comorbid major depression show significantly greater impairment in health-related quality of life than in either disorder alone. Primary care patients with GAD showed significantly higher annual medical costs than patients without GAD (median $2,375 versus $1,448) and higher mean annual medical costs ($2,138) than patients with other anxiety disorders. GAD is frequently under-recognized in primary care, and only 20% to 32% of patients receive adequate treatment. Suboptimal treatment adds to the health-related quality of life burden of this disorder  .