To the Editor: We thank Dr. Capuron and colleagues for their interest in our article. They object to our comparison of cytokines to hormones. This comparison is not new and has often been made to underscore both the similarities and the differences in their regulation and mechanism of action (1). One major difference with regard to regulation is that cytokines are usually secreted on demand in massive amounts relative to their baseline levels, which are often undetectable. Cytokine secretion puts into motion a series of events, most of which are initiated by binding to a specific cytokine receptor, and in this regard, cytokines are similar to hormones. A significant difference between cytokines and hormones, however, is that cytokine-induced cell activation often causes the release of soluble receptors, sometimes in concentrations a hundred times higher than the specific cytokine released (2). The purpose of this tight regulation of cytokine secretion and action may be to initiate an immune and/or inflammatory response at the local level while protecting the rest of the organism from the potentially devastating effects of unchecked levels of cytokines (3). We therefore agree that differences in circulating levels of any specific cytokine between two subject groups are unlikely to provide useful information about cytokine regulation in these subjects. This was explicitly stated in our article. For this reason, we limited our review of the literature on human stress and cytokines to in vitro cytokine production by stimulated leukocytes.

As to the issue of cytokine regulation in psychiatric patients, we were careful to emphasize in our article only studies in which cytokine secretion or other indices of cytokine physiology were measured directly. Indirect evidence for a possible role for cytokines in major depression—no matter how appealing—remains more in the realm of speculation and theory awaiting confirmation than as reproducible fact. Also, we think it is premature to link evidence of cytokine dysregulation in schizophrenia to brain swelling or disruption of the blood-brain barrier, since direct evidence for these observations remains slim. Last, many "omissions" referred to in the letter pertain to publications that have appeared since the submission of our manuscript. We would welcome the opportunity to add them to our overview.