Once someone has had a "major depressive episode" (using criteria specifically described in the DSM-IV), future serious depressive episodes are often described as being a recurrence of "major depressive disorder." A "major depressive episode" is not just a "depression"; a look at the criteria for diagnosing it will indicate that it's a pretty serious occurrence. Once "major depressive disorder" is appropriately diagnosed, the symptoms may at times be severe, moderate, mild, or in remission. The general criteria are in the DSM. Whether or not any improvement is stable (with or without treatment) is an important issue for safety and risk assessment (e.g., related to danger of suicide).

There are other kinds of depression that deserve treatment, some of which should involve consideration of medication or other biological treatments. A person with major depressive disorder may separately meet other mood disorder diagnostic criteria (including criteria for other depressive disorders), and may experience milder depressions unrelated to the pre-existing mood disorder (just as anyone else). For example, "dysthymia" (once again, as defined in the DSM, not some informal definition), is a condition that is painful and chronic but in which the person/patient never has a true "major depressive episode." It can occur by itself or along with another mood disorder.

As you know, we're not talking about brief bouts of blues, expected grieving, or expected reactions to situations such as losing a job, school/money/relationship stresses or "life problems." One can become seriously depressed in connection with situational events, but there is usually a qualitative difference between a person who develops (or has a recurrence of) major depressive disorder or dysthymia and one who has some sort of acute stress disorder or adjustment disorder. Each has its challenges and opportunities for successful treatment.

RE: Bipolar disorder: I agree that it is probably overdiagnosed in a significant number of people. In my view, the initial diagnosis should generally involve a review of reliable history (history from the patient alone may or may not be sufficiently reliable), an adequate examination/diagnostic interview, and careful follow-up to be sure that continuing symptoms, signs and treatment response are consistent with the diagnosis. Once diagnosed, the concept stays with the patient for life, and can be misunderstood by both the patient and others (especially if the diagnosis is iffy). Nevertheless, prompt diagnosis creates an opportunity for proper treatment; I would not want such a serious condition to be particularly underdiagnosed, either. I see too much tragedy associated with underdiagnosis, inadequate treatment, and insufficient follow-up of serious mood disorders.