Corticosteroids (also known as corticoids) are hormones secreted by the outer portions of the adrenal glands. They can be divided into three groups, for each of which there are separate receptors: androgens, which are involved in the development of sexual traits; mineralocorticoids (aldosterone, corticosterone, desoxycortisone), which regulate the body's osmotic balance; and glucocorticoids (cortisone, hydrocortisone, prednisone), which, in addition to their anti-inflammatory and immunosuppressive effects, stimulate the synthesis of glucose and increase the mobilization of fatty acids and proteins to meet the higher metabolic demands generated by stress.

Glucocorticoids play an extremely important role in fear and anxiety reactions and in depressive states. These hormones often affect behaviour by increasing or decreasing the efficiency of certain neural pathways.

It has long been known that depressed persons display hyperactivity in the hypothalamic-pituitary-adrenal (HPA) axis (see illustration and explanation above). A prolonged state of inhibition of action is also known to encourage the emergence of a depressive state. This chronic excess stress on the HPA axis is believed to result in structural changes in certain parts of the brain. For example, region CA3 of the hippocampus loses large numbers of neurons when subjected to prolonged stress.

Other studies have also reported a reduced number of glucocorticoid receptors in the hippocampus and prefrontal cortex of suicide victims. Though it is hard to tell whether these structural changes are of genetic origin or the result of chronic activation of the HPA axis, they would be consistent with hyperactivity in this axis when the natural braking effect of these receptors was reduced.

Here's another example: people with Cushing's syndrome, a disease in which the body produces excess cortisol, display a high incidence of depression, and their depression lifts when they are given treatments that bring their cortisol levels back to normal.

Thus, all indications are that the end products of the HPA axis—glucocorticoids— play a role in depression by influencing several neurotransmitter systems, including those for serotonin, norepinephrine, and dopamine, all three of which are involved in depression.

Treatment with antidepressants is often regarded as consisting of two phases. During the first two weeks, the patient's depressive state does not really improve. But after two or three weeks, the patient gradually begins sleeping and eating again, feeling more energetic, and having more positive thoughts. It is recommended that treatment then be continued for several months to minimize the risk of a relapse.

Various hypotheses have been offered to explain this lag before antidepressants become effective. One hypothesis is that at the start of the treatment, once the antidepressant medication has inhibited the reuptake of serotonin, the serotonin autoreceptors quickly become saturated, so that their inhibitory effect predominates, reducing the amount of serotonin released into the synaptic gap. But over time, these autoreceptors become desensitized, and the presynaptic neuron can produce action potentials more readily. Because the antidepressants are still preventing the serotonin from being reabsorbed, its extracellular concentration increases, and serotonergic transmission is facilitated.

Another interesting hypothesis deals less with the efficiency of existing connections that with the development of completely new neurons.

The effect of antidepressants can be compared with that of ecstasy, which causes the release of large amounts of serotonin from the nerve endings of neurons. This excess of serotonin is suspected to be the source of the particular mental effects of ecstasy, including those associated with feelings of well-being—effects analogous to those of antidepressants.

Many researchers now believe that it is not really appropriate to describe the physiological causes of depression as a "chemical imbalance".

The hypothesis that depression was caused by a "chemical imbalance" originated in the 1960s. The first antidepressant medications, developed around that time, were tricyclics and monoamine oxidase inhibitors (MAOIs). In addition to alleviating the symptoms of depression in many patients, these molecules were known to increase brain levels of dopamine, norepinephrine, and serotonin in various ways. For this reason, researchers hypothesized that depression might be due to an imbalance in these neurotransmitters. This hypothesis did indeed yield some fairly useful research findings during the last decades of the 20th century. In addition, by emphasizing that mood disorders might be related to a physiological malfunction and not simply to a character defect or a lack of will power, this hypothesis reduced the needless feelings of guilt that often haunted people with depression.

However, the results of efforts to identify this "chemical imbalance" more precisely have been rather disappointing and contradictory. Research is now focusing less on the neurotransmitters themselves and more on the receptors for these molecules and on the molecular events involved in regulating genes. But here too, there is room for controversy. Relatively little direct evidence has been found of alterations in receptors or anomalies in gene expression related to these receptors or other enzymes in cases of depression. Moreover, the reason for the two to three week lag between the time when antidepresssants first affect neurotransmitters and the time when they begin to affect mood (see sidebar to the left) is still not well understood. In short, the situation is far more complex than scientists believed in the 1960s when they first formulated the "chemical imbalance" hypothesis.

Given these problems in securing any unequivocal data to support this hypothesis, some scientists have begun to ask whether the extensive use that continues to be made of the term "chemical imbalance" might raise some ethical problems, or even political ones. In the United States, for example, where advertisements for antidepressants are allowed in the mass media, the pharmaceutical companies have not always worried about this fine point. Simplistic advertisements tell Americans that when they are depressed, a substance in their brain is out of balance, and that if they take the right antidepressants, the ideal balance will be magically restored. These advertisements may well have something to do with the stunning success of SSRI antidepressants, such as Prozac, Zoloft, and Paxil, in the marketplace, and the billions of dollars that they have earned for the companies that make them.

The effects of antidepressants are not limited to the presynaptic neurons. In postsynaptic neurons, the antidepressive effect of tricyclics and MAOIs may be attributable to “down-regulation” (reduction in the number but not the sensitivity) of the beta-adrenergic receptors and the 5-HT₂ serotonergic receptors. Desensitization is also observed in the norepinephrinergic receptors coupled to adenylate cyclase . The phenomena of transduction via the G-proteins paired with the receptors represent another possible site of postsynaptic effects, as is probably the case for lithium.