That's an old topic to discuss, back again because of yet another study which finds that anti-depressants may not be effective for all levels of depression:
The latest study may settle a debate about drugs like Prozac.
The findings could help settle a longstanding debate about antidepressants. While the study does not imply that the drugs are worthless for anyone with moderate to serious depression — many such people do seem to benefit — it does provide one likely explanation for the sharp disagreement among experts about the drugs' overall effectiveness.
Taken together, previous studies have painted a confusing picture. On one hand, industry-supported trials have generally found that the drugs sharply reduce symptoms. On the other, many studies that were not initially published, or were buried, showed no significant benefits compared with placebos.
The new report, appearing in The Journal of the American Medical Association, reviews data from previous trials on two types of drugs and finds that their effectiveness varies according to the severity of the depression being treated.
...
Three of the trials were of Paxil, from GlaxoSmithKline, a so-called S.S.R.I., and the other three were of imipramine, an older generic drug from the class known as tricyclics. The team, led by Jay C. Fournier and Robert J. DeRubeis of the University of Pennsylvania, found that compared with placebos, the drugs caused a much steeper reduction in symptoms of severe depression (cases scoring 25 or higher on a standard scale of severity, putting them in the top quarter of the sample). Patients with scores of less than 25 got little or no added benefit from the medications.
My first reaction to reading this was to wonder if mild depression indeed is just a milder form of the same biological condition as more severe depression. The results could follow if the two are somehow different conditions. Many other explanations are also possible.