Abstract
Establishing the optimum therapeutic dose of a putative antidepressant is important to avoid unnecessarily high doses, which may be associated with increased frequency of severity of unwanted side effects, or too low a dose which may not achieve the best therapeutic effect. Flexible dose regimes that are often used in clinical trial programmes may lead to the use of too high a dose because of the attribution of response in depression to the higher dose used later in the study rather than being identified as a delayed response to a lower dose used earlier. Fixed dose studies provide a more reliable view of the dose response relationship and, where differences between doses are small, metanalysis of large databases may provide a useful tool for the establishing of the minimum therapeutic dose. Metanalysis of the placebo controlled results with citalopram demonstrated that the dose-response curves based on log odds ratios showed a very flat curve across the 20-60 mg range and that 20 mg appeared therefore to be the minimum effective dose. There was evidence that in some subgroups of depressed patients a better response may be seen with a higher dose. For example in patients with severe depression citalopram was effective compared with placebo in doses of both 20 mg and 40 mg. However there was a more pronounced therapeutic effect with the higher dose. Similar results have been reported with other selective serotonin reuptake inhibitors.